37.35

1
.1

 

 

5'
1;}!

t!
‘V

 

. i
v!) I
:4

9.1113,!

at: 1...:

film
a .
An ".3

 

i
u

U

.‘~illll\-v§

H

 

.fiarfl Ea. “my;

 

 

C‘ - r

i ll

06%)

MICHIGAN STA RSITY LIBRARIE

l NWillllliml’lllill/Hillllllll

3 1293 01388 3289

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

Risk Assessment of Transgenic Plants:
Evaluation of Border Rows as a Containment Strategy
for Transgenic Pollen and a Comparison of Pollen
Dispersal Patterns for Native and Transgenes

presented by
Stan C. Hokanson

has been accepted towards fulfillment.
of the requirements for

PhD

 

degree in We
Plant Breeding and
Genetics Program
Ecology and Evolutionary
Biology Program

(lg-M / 2‘6 {:fi‘vfl’ “9%““’3//

F Major professor

Dam 12121/95

MSUt's an Affirmative Action /Equal Opportunity Institution 0- 12771

 

LIBRARY
Michigan State
University

 

 

 

PLACE u RETURN BOX to romwothb mum you} mead. '
TO AVOID FINES Mum on or baton date duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU IsAnNflnnatm Action/Emil Oppommltylmtttulon
W1

RISK ASSESSMENT OF TRANSGENIC PLANTS:
EVALUATION OF BORDER ROWS AS A CONTAINMENT
STRATEGY FOR TRANSGENIC POLLEN AND A COMPARISON
OF POLLEN DISPERSAL PATTERNS FOR
NATIVE AND TRANSGENES

By

Stan C. Hokanson

A DISSERTATION

Submitted to -
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Horticulture
Plant Breeding and Genetics Program
Ecology and Evolutionary Biology Program

I995

ABSTRACT
RISK ASSESSMENT OF TRANSGENIC PLANTS:
EVALUATION OF BORDER ROWS AS A CONTAINMENT
STRATEGY FOR TRANSGENIC POLLEN AND A COMPARISON
OF POLLEN DISPERSAL PATTERNS FOR
NATIVE AND TRANSGENES
By

Stan C. Hokanson

Despite full commercial approval of twelve transgenic crops in the US. (circa
1995), concern is still being expressed regarding the potential risks associated with the
agronomic-scale production of transgenic crops. One commonly mentioned concern
involves the pollen-mediated escape of engineered genes into populations of crop wild
relatives. In this study two questions relevant to this issue were investigated: 1) Can
plantings of border rows effectively limit pollen mediated gene movement, and 2) Do
the pollen-mediated dispersal patterns of transgenes differ from those of native genes?
The ratio of recessive trap plants to wild type donor plants was varied to test the
efficacy of border rows as a means to limit the spread of transgenic pollen to
discontiguous satellite plots. Gene movement within the border plots assumed a
leptokurtic distribution. Increasing the number of donor plants increased levels of
gene flow both within the border and to the discontiguous satellite plots. As the
trap/donor ratio increased, there was a significant decrease in long distance gene
movement to the satellites, although the observed year to year and site to site
variability could limit the effectiveness of this strategy. Furthermore, extremely large

numbers of border plants would be required to minimize pollen movement on a

commercial scale. Dispersal patterns of transgenes and native genes were evaluated by
comparing levels of pollen-mediated gene movement from melon plants (Cucumis
melo) expressing dominant morphological and transgenic marker genes into a
surrounding border of recessive non-transgenic melon plants. Long distance dispersal
patterns for the two genes were identical and dispersal patterns into the plot borders
were nearly identical. Several of the apparent discrepancies were explained by
transgene inactivation, a phenomenon which has implications for any study measuring
gene movement with transgenic plants. Results from this study validate the
assumption that native and transgenes have the same dispersal patterns. Thus,
application of non-transgenic results to transgene escape and dispersal issues should be
appropriate. However, the assessment of establishment and spread will depend on
both pollen movement and the fitness value of the particular transgene crop

combination.

ACKNOWLEDGEMENTS

I was born in the age of rockets and grew up on a dirt road between two small
towns. In large part this seemingly insignificant juxtaposition of two worlds has led
me here to the point of completing this project. Perhaps what I mean to say is,
finishing this project represents the culmination of a long journey. However, the task
at hand is not to explain the philosophical basis of my scientific life (which I probably
could not accomplish anyway). Rather, my intent here is to attempt to acknowledge
the people who have made this chapter of my life possible. To name everyone would
be nearly impossible and yet at some risk I will name a few. The mere mention of
these names in no measure expresses my gratitude. In the same vein, my failure to
mention here all who played a role in this chapter of my life does not diminish their
role. If you sweated, fretted, laughed or cried with me along the way, the depth of
that emotional bonding is the true measure of my gratitude

Early on several people strongly encouraged me to go back to school. My
mother has always been my staunchest advocate. Her unwavering faith in me helped
start this ball rolling. My good friends Dick and Deb Rieth and Ken Kirton have

never failed to love and encourage me as I expect they never will.

iv

Once in graduate school the friendships came fast and furiously with all the
help and encouragement one could ever hope for. My friends Pete and Mollie Callow
offered me a place to stay and warm meals when I was but a name on paper. That
early act of kindness has blossomed into a fast friendship. From the beginning Steve
Krebs took me under his professional wing, providing both practical advice and very
enjoyable eclectic discussion. He and his wife Roberta have become a constant source
of friendship and encouragement. Roger and Lori May have shared the trials and
tribulations of this graduate student life cheerfully and generously. The comradery,
help and encouragement they provided was boundless. At times, Sue Hammar all but
carried me in the lab. A similar role was played by Dan Prince and John Holmes in
the field and Camille Ciesliga in the greenhouse.

I thank my committee members both past and present; Steve Tensor, Steve
Malcolm, Bernie Zandstra, and Andy Jarosz for the thoughtful reviews of my
manuscripts and dissertation and their non-threatening bearing which made this process
more enjoyable and rewarding.

My advisors Rebecca Grumet and Jim Hancock demonstrated an extraordinary
concern for my progress as a graduate student. Rebecca's precision thinking and
writing skills served me well at all times. However, her desire to excel in all aspects
of the student/advisor process was an inspiration for me.

Jim Hancock took me into his lab when all I could do was hoe grapes and
speak vaguely of my aspirations. He has delivered me to the threshold of those
aSpirations. Over the years Jim has served in all the typical advisor roles and some

not so typical. He has been a dependable training partner and a wonderful friend. I

have seen him take the risk of becoming involved in the personal/emotional aspects of
people's lives, delivering a human element to this process we call higher education.
Jim's natural tendency to see the good in the people and situations around him gives
him the ability to draw out the best in those around him without intimidating or
pressuring. Jim also has an uncanny ability to clearly see the fulcrum on which issues
reside. This ability has simplified many issues for me over the years. I want to thank
Jim for providing a non—threatening environment in which I could learn how I wanted
to do science and find my new path in this world!

Last I thank my wife Karen and daughter Kate who have given me a whole
life. From them I have learned lessons I never expected to learn in this institution of
higher learning. Karen has become my best friend in this world. Her ability to carry
so many tasks forward with such quiet efficiency and grace has won my undying

admiration. Her spirit has won my undying love.

vi

Guidance Committee:

The journal paper format was chosen for this thesis in accordance with
departmental and university regulation. The thesis is divided into three chapters and
two appendices. Chapter 1 is in press in HortScience. Chapter 2 has been submitted
for review to Ecological Applications. Chapter 3 will be submitted to Theoretical and
Applied Genetics. Appendix 1 and 2 are intended as a guide for future herbicide

selection experiments planned as an extension of the work presented herein.

vii

TABLE OF CONTENTS

Page
LIST OF TABLES .......................................................................................................... x
LIST OF FIGURES ........................................................................................................... xii
INTRODUCTION ........................................................................................................... 1
CHAPTER 1: CHARACTERIZATION OF THE BLUNT LEAF APEX (bla )
TRAIT IN CUCUMBER ................................................................................................... 16
ABSTRACT ........................................................................................................... 17
LITERATURE CITED .......................................................................................... 25
CHAPTER 2: EFFECT OF BORDER ROWS AND TRAP/DONOR RATIOS
ON POLLEN-MEDIATED GENE MOVEMENT .......................................................... 26
ABSTRACT ........................................................................................................... 27
INTRODUCTION ................................................................................................. 28
MATERIALS AND METHODS ......................................................................... 32
RESULTS .............................................................................................................. 38
DISCUSSION ........................................................................................................ 48
LITERATURE CITED .......................................................................................... 51
CHAPTER 3: COMPARISON OF DISPERSAL PATTERNS OF A
TRANSGENE AND NATIVE MARKER GENE ORIGINATING FROM
THE SAME DONOR ....................................................................................................... 57
ABSTRACT ........................................................................................................... 58
INTRODUCTION ................................................................................................. 59
MATERIALS AND METHODS ......................................................................... 61
RESULTS AND DISCUSSION .......................................................................... 71
LITERATURE CITED ......................................................................................... 83

viii

TABLE OF CONTENTS (continued)

Page
CONCLUSIONS ................................................................................................................ 89
APPENDIX 1 ..................................................................................................................... 95
APPENDIX 2 ..................................................................................................................... 96

ix

LIST OF TABLES

Table Page
Chapter _1_
1. Test for co-segregation of blunt leaf apex and cordate leaf base ............................. 22

2. Comparison of flowering and fruiting of ’Wisconsin SMR-l8' and 'Wisconsin
SMR-18' bla genotypes ................................................................................................. 24

QM;

1. Trap/donor ratio for each of the four treatments with sites and locations for
the replications of each in 1992 and 1993 .................................................................. 36

2. Mean long distance gene movement to the satellite plots for the four
treatments in 1992 and 1993 ........................................................................................ 46

3. Long distance gene movement (expressed as the % wild type seedlings in
each satellite) to the eight individual satellites within each of the twelve
plots tested in the summers of 1992 and 1993 ........................................................... 47

LIST OF TABLES (continued)

Table Page
Client‘s—3
1. Test for independent segregation of virescence (vir) and neomycin
phosphotransferase (NPT II) ......................................................................................... 63
2. Trap/donor ratios and locations for each of the treatments tested in 1994 ............. 66

3. Long distance movement for the native and transgene to the satellite plots
for the four sites in 1994 ............................................................................................... 72

4. Phenotypic combinations for progeny of recipient, non-transgenic,virescent
plants. Presence of NPT II was analyzed by ELISA ................................................ 75

LIST OF FIGURES
Figures Page

Chapter _1_

l. 310 Phenotypes. A. Variation in leaf shape for first true leaves of
'Wisconsin SMR-l8' bIa. Arrows indicate a 'Wisconsin SMR-18' bla
individual with a first leaf with an acute apex and truncate leaf base.
B. Leaf phenotypes for 'Wisconsin SMR-l8' (left) and 'Wisconsin
SMR-18' bIa (right). Note the indented or cordate leaf base for the
'Wisconsin SMR-18' leaf on the left and the truncate base for
'Wisconsin SMR-18' bid on the right. C. Progeny from an open-
pollinated 'Wisconsin SMR-18' bIa plant. White dots indicate wild
type 'Wisconsin SMR-18' leaves originating from a heterozygous
seedling (an outcrossing event). Note other leaves have the recessive
truncate leaf base .......................................................................................................... 20

Chapter 2

l. (A) Overall field plot design used in 1992 and 1993 experiments. All 12
plots regardless of design were surrounded by eight satellite plots 50
meters distant, each containing four recessive SMR-18 bla plants.

(B-E) Four treatments tested in these experiments depicted from highest

to lowest trap to donor ratio. (B) 1m2 of SMR-18 donors surrounded

by a 400m2 border of recessive SMR—l8 bla plants. (C) 1m2 of donors

surrounded by a 100m2 border of recessive recipients. (D) 4m2 of donors
surrounded by a 100m2 border of recessive recipients. (E) lm2 of SMR-18

donors with no surrounding border plants (borderless). At harvest

borders were subdivided into 1m2 subplots and 2-3 fruit were collected

from each subplot (see also Fig. 3). A Location of bee hives on the

edge of the donor plots ................................................................................................. 34

xii

LIST OF FIGURES (continued)

Figure Page

2.

Comparison of short distance gene movement (percent gene movement

observed within the borders at one meter intervals from the plot center)

for plots with lmz, A, or 4m2, -, donor plots. One square meter data are

the mean of all plots with 1m2 donors in 1992 and 1993. 4m2 data are the

mean of the two 4m2/100m2 plots in 1992. Data for paired comparisons

came from 1m2 and 4m2 donor plots with 100m2 borders at two locations

in 1992 ............................................................................................................................ 4O

. Representative example of gene movement into the border subplots

(plotted as % wild type seedlings among all seedlings in each 1m2
subplot) for the 1992 Kell-l plot. Each square represents a 1m2
subplot ............................................................................................................................. 42

. Mean values (i standard errors) for long distance gene movement

(% of total seedlings with the dominant marker trait) to satellite plots for
the four treatments tested in 1992 and 1993, plotted as a function of the
trap/donor ratio for the treatment ................................................................................. 45

Chaptar 3

. A. Overall field plot design. Each main plot (center square), was

surrounded by eight satellite plots (shaded), located 50 meters from the

plot center. Each satellite plot contained four non-transgenic

recessive virescent melon plants. B and C. Expanded view of center

plots. B. The bordered plot contained 1m2 of transgenic non-virescent

donor plants (small center square), surrounded by a 100m2 border of
non-transgenic recessive virescent melon plants. C. Borderless center

plot contained 1m2 of transgenic non-virescent donor plants.

A Location of bee hive on the edge of the donor plots ............................................. 65

. Percent gene movement for the native (Vir), O, and the transgene

(NPT II), A, plotted as a function of distance in meters from the donor
plot center. Gene movement is calculated as the percent Vir (native) and
percent NPT II (trans) positive among all seedlings scored ...................................... 74

xiii

LIST OF FIGURES (continued)
Figure Page

3. Plot map depicting the Sandhill border subplots (numbered 1-132), and
the location of the fruits which produced the 39 NPT H ELISA‘ green
seedlings (shaded subplots). The six striped subplots in the center
depict the location of the donor plants. Numbers in the shaded subplots
represent; (Top), total number of green NPT II ELISA' seedlings
originating from the subplot, (Bottom), total number of NPT II ELISA‘
seedlings found to contain an NPT II gene with PCR .............................................. 77

4. PCR products resulting from a reaction run with the two 18 base pair
primers flanking a 700 base pair region internal to the NPT II gene, and
genomic DNA from: Lanes 2-3, NPT H ELISA+ green seedlings; Lane 4,
NPT II ELISA‘ virescent seedling; and Lanes 5-10, NPT II ELISA‘ green
seedlings. Band in lanes 2-3 and 5-10 corresponds to approximately 700
base pairs based on comparison with the Hind III digested A DNA size
standard, Lane 1 ............................................................................................................. 80

xiv

INTRODUCTION

The first breakthroughs in genetic engineering technology occurred in the early
1970's. Included among the many breakthroughs of this era were the ability to clone
DNA in a bacterial host or vector (1973), the discovery of restriction enzymes capable
of cutting DNA at specific sites (1975) and the ability to determine the exact base
sequence of a DNA fragment (1975). In 1983 researchers at the university of Ghent
and at the Monsanto Co. had independently uncoupled the crown gall causing genes
from the Ti plasmid of Agrobacterium tumefaciens (Zambryski et al., 1983; Fraley et
al., 1983). By replacing the tumor inducing region with a DNA sequence of interest
they had created a bacterium capable of transfering foreign DNA into the plant
genome and thus was born the era of plant genetic engineering. Since that time
progress in the develOpment and testing of transgenic plants has been dramatic. Today
approximately 60 plant species have been or are currently in the process of being
genetically engineered (Raybould and Gray, 1993; Rissler and Mellon, 1993; Rissler
and Mellon, 1994; Rogers and Parks, 1995). Moreover, twelve transgenic crop/gene
combinations have now been given full commercial approval (Rissler and Mellon,
1995). It has been suggested that genetic engineering will become the fourth wave in
modern food production following selective breeding, modern hybrids and the use of

petrochemicals (Gasser and Fraley, 1989; Fraley, 1992). However, commensurate with

2

fast development of the technology, and the excitement generated by the potential
benefits to be provided by it, concern has been voiced regarding the potential risks
associated with the production and consumption of genetically engineered crops on a
commercial scale.

These concerns fall into three general categories; 1) the potential risk posed by
transgenic crops to human health, 2) the potential risk posed to the domesticated plants
and animals on which we depend, and 3) the potential risk that is posed to the natural
environment by genetically engineered crops. There are a number of potentially
negative interactions which could occur between genetically engineered mom and
natural ecosystems. The research described within this dissertation addresses the
general area of the potential risk of pollen-mediated escape of transgenes from crop
plantings into natural populations of crop wild relatives.

Gene flow has long been recognized as an elemental evolutionary force.
Migration of a few individuals between small populations can offset the effects of
random drift and selective pressures (Wright, 1951; Antonovics, 1968). Much effort
has been directed toward understanding the mechanisms and implications of gene flow
in both a theoretical and empirical sense. Along these lines, one of the most
prominently mentioned risks associated with the commercialization of transgenic mom
has been the potential for pollen-mediated escape of engineered genes (Colwell et al.,
1985; Ellstrand, 1988; Ellstrand and Hoffman, 1990; Dale, 1992; Raybould and Gray,
1994)

If established in natural populations, transgenes could effect population

dynamics within the species complex itself or at the community level. At the species

3

level a transgene conferring a high fitness value to it's carrier could cause a major shift
in allele frequencies within the population raising the possibility of the loss of rare
alleles not associated with the transgene (Regal, 1994). Even in the absence of a
selective advantage, "swamping" of small natural populations with domesticated genes
from large agronomic plantings could lead to the extinction of many ”wild" genes
(Ellstrand, 1992), or species (Small, 1984). Species loss could include crop wild
relatives in the centers of crop diversity (Rissler and Mellon, 1993). These wild
relatives presumably contain, among other things, potentially valuable genes for future
disease or pest resistance. Erosion of genetic diversity in centers of diversity is
already occurring at an alarming rate (Fowler and Mooney, 1990).

At the community level, transgenes which improve the plant’s fitness could lead
to an ecological release (Schmitt and Linder, 1994). Such a release could lead to an
expansion of the species within the community or expansion into a new ecological
range. These expansions have the potential of mirroring invasions of non-native
introduced pests (Mooney and Drake, 1986; 1990). Invasions of non-native pests have
caused among other things, species displacement, interruption and redirection of
successional change, and changes in abiotic conditions such as moisture and salinity
levels and soil nutrient and biotic composition. Changes such as these have the
potential of fomenting a cascade of similar changes in an ecosystem.

It is now quite evident that the risk of engineered genes escaping into natural
populations is real. Of the worlds 20 major crop species, all, with the exception of
soybean, Glycine max, peanut, Arachis hypogea, coffee, Coflea ambica, chick pea,

Cicer arietr'num, and sweet potato, Ipomoea batatas, have been found to naturally

4

produce hybrids with their wild relatives (Hancock et al. manuscript in prep; Ellstrand
et al., manuscript in prep.)

Whether or not transgenic crops will be grown in close enough proximity to
their wild relatives to allow pollinations/hybridizations to occur is probably also a
foregone conclusion. Although there are only ten crop species with compatible wild
relatives in the US, compatible wild relatives of all our crop plants are found
somewhere (Hancock, et al., manuscript in prep). It is unlikely that the use of
transgenic crops will be restricted to areas where the crop and its wild relatives do not
come in contact. In some of the areas of the world where the perceived need for
transgenics is highest they are most likely to come in contact with wild relatives
(Hodgson, 1992; Gershon, 1992; Miller et al., 1995). In fact, transgenic crops are
being commercialized most aggresively in just some of these regions (Moffat, 1994).
Examples include Mexico, the center of diversity for maize, Central and South
America, center of diversity for crops such as squash, potato, tomato and peanut, and
Southeast Asia, center of diversity for rice, banana, citrus and sugar cane.

Given the near certainty that domesticated genes will escape if fertile
transgenics are planted in close proximity to their wild congeners, several mechanisms
for containing the pollen-mediated spread of transgenes to their wild relatives have
been preposed. These include: 1) Isolating the transgenic crop by distance from its
wild relatives, 2) Using barrier or guard rows to trap or intercept transgenic pollen
from leaving the plot, or 3) Genetically isolating the crop through the use of male
sterility or pollen-lethal genes (Ellstrand, 1988; Ellstrand and Hoffman, 1990; Kareiva

et al., 1994). Tests of these mechanisms have been performed; isolation by distance,

5

(Manasse, 1992; Morris et al., 1994), barrier or guard rows, (Tynan et al., 1990;
Umbeck et al., 1991; Scheffler et al., 1993; Morris et al., 1994), male sterility, (Eber et
al., 1994). Results from the tests of the isolation by distance mechanism suggest that
barren zones or increased distance between blocks of plants might actually serve to
increase the amount of gene movement out‘ of the isolated plot rather than decrease it.
Studies of the extent of gene movement in border plantings uniformly demonstrate that
most pollen is deposited within a few meters of the source. However, none of these
studies was designed to measure pollen movement beyond a contiguous border
planting. Many such studies report a clustering of donor genes on the border edges
suggesting that donor pollen may be moving beyond the border edge. Finally, in an
experiment designed to evaluate the level of outcrossing of rapeseed, Brassica napus,
to weedy relatives and the purity of F,s produced in the presence of the weedy
relatives when the cultivar was male sterile, it was found that spontaneous interspecific
hybrids could be produced under natural conditions when the male sterile served as the
female parent (Eber et al., 1994).

The specific goals of this dissertation were to address the following questions:
1) Can border rows be used as a means to effectively limit pollen movement, and 2)
Are the pollen dispersal patterns of native and transgenes the same?

The initial chapter of the dissertation is a further characterization of the
cucumber (Cucumis sativus L.) mutant, 'Wisconsin SMR-l8 bla' (blunt leaf apex)
facilitating its use as an isogenic recipient parent to study pollen movement. A new
character, (truncate leaf base) associated with previously described characters provided

a quick, reliable screen for the large numbers of seedlings evaluated in the experiments

6

described in Chapter two. Chapter two of this dissertation describes a set of
experiments designed to test the efficacy of border rows as a means to restrict the
pollen-mediated movement of transgenes out of mm plantings. The number of donor
plants was varied in conjunction with varying border sizes to create four trap plant to
donor plant ratios. The influence of these ratios on long distance gene movement to
discontiguous satellite plots located 50 meters from the plot centers was evaluated.
Although all the aforementioned studies, including our own, reported varying degrees
of success in restricting the escape of engineered genes, protection was never
complete. Taking into account the results of such tests, the consensus seems to be that
regardless of the containment strategy employed, some genes will escape from
agronomic scale plantings of transgenic crops.

With the knowledge that transgenes will escape, the next level of concern
revolves around the nature of gene movement itself. What is the likelihood that
transgenes will become established in natural populations? And what will be the rate
of spread of these genes once they become established? Research dealing with actual
nansgene movement and establishment is only beginning to emerge (Crawley et al.,
1993; McPartlan and Dale, 1994; Scheffler et al., 1993; Scheffler and Dale, 1994;
Linder and Schmitt, 1994, 1995). Due to this limited amount of information
generated by research with actual transgenic plants, much of the response to these
issues has been based on a body of theoretical and empirical evidence accumulated
from research done with non transgenic organisms (Andow, 1994; Crawley,
1987,1990; Darmency, 1994; Gliddon, 1994; Manasse and Kareiva, 1991; Mooney and

Drake, 1990; Williamson, 1994).

7

The widespread use of this non-transgenic data base to respond to transgenic
risk assessment issues is based on the assumption that transgenes will disperse in the
same fashion as native genes. While for many this seems to be a reasonable
assumption, the novel sources of some transgenes (Rissler and Mellon, 1993) and the
fact that transgenes are not always predictably expressed in plants (Finnegan and
McElroy, 1994) lead others to believe that any assumption concerning genetically
engineered crops should be evaluated (Rissler and Mellon, 1993). Chapter three of
this dissertation details a direct comparison of dispersal patterns for a transgene and
native gene originating from the same donor. Additionally, we revisit the influence of

border rows on long distance movement for the two types of genes.

8

Literature Cited
Andow, DA. 1994. Community response to transgenic plant release: Using

mathematical theory to predict effects of transgenic plants. Molecular Ecology

3:65-70.

Antonovics, J. 1968. Evolution in closely adjacent populations. VI. Manifold effects

of gene flow. Heredity 23:507-524.

Colwell, R.K., EA. Norse, D. Pimentel, F .E. Sharples and D. Simberloff 1985.

Genetic engineering in agriculture. Science 229:111-112.

Crawley, M.J. 1987. What makes a community invasible? pgs. 429-453 in
Colonization, Succession and Stability [eds] A.J. Gray, M.J. Crawley and R].

Edwards. Blackwell Scientific Publications, Oxford

Crawley, M.J. 1990. the ecology of genetically engineered organisms: Assessing the
environmental risks. pgs. 133-150 in Introduction of Genetically Modified
Organisms into the Environment [eds] HA. Mooney and G. Bemardi. Scope

44 John Wiley & Sons, Chichester.

Crawley, M.J., RS. Hails, M. Rees, D. Kohn, and J. Buxton 1993. Ecology of

transgenic oilseed rape in natural habitats. Nature 362:620-623.

9

Dale, RI. 1992. Spread of engineered genes to wild relatives. Plant Physiology

100:13-15.

Darmency, H. 1994. The impact of hybrids between genetically modified crop plants

and their related species: introgression and weediness. Molecular Ecology

3:37-40.

Eber, F., AM. Chévre, A. Baranger, P. Vallée, X. Tanguy and M. Renard 1994.
Spontaneous hybridization between a male-sterile oilseed rape and two weeds.

Theoretical and Applied Genetics 88:362-368.

Ellstrand, NC. 1988. Pollen as a vehicle for the escape of engineered genes? Pgs.
330-332 in J. Hodgson and AM. Sugden [eds] Planned Release of Genetically
Engineered Organisms. Trends in Biotechnology/Trends in Ecology and

Evolution Special Publication. Elsevier Publications, Cambridge UK.

. 1992. Gene flow by pollen: implications for plant conservation genetics.

Oikos 63:77-86.

Ellstrand, NC. and CA. Hoffman 1990. Hybridization as an avenue of escape for

engineered genes. BioScience 40:438-442.

10

Finnegan, J. and D. McElroy 1994. Transgene inactivation: Plants fight back!

BioTechnology 12:883-888.

Fowler, C. and P. Mooney 1990. Shattering: Food, Politics, and the Loss of Genetic

Diversity. University of Arizona Press. Tucson, AZ.

Fraley, R.T., S.G. Rogers, RB. Horsch, PR. Sanders, J.S. Flick et al., 1983.

Expression of bacterial genes in plant cells. Proceedings of the National

Academy of Sciences, USA. 80:4803-4807.

Fraley, R. 1992. Sustaining the supply. Bio/Technology 10:40-43.

Gasser, CS. and RT. Fraley 1989. Genetically engineering plants for crop

improvement. Science 244: 1293-1299.

Gershon, D. 1992. Programme aids developing world. Nature 356:735.

Gliddon, C. 1994. The impact of hybrids between genetically modified crop plants

and their related species: biological models and theoretical perspectives.

Molecular Ecology 3:41-44.

Hodgson, J. 1992. BioTechnology: Feeding the world? Bio/Technology 10:47-51.

11
Kareiva, P., W.F. Morris and CM. Jacobi 1994. Studying and managing the risk of

cross-fertilization between transgenic crops and wild relatives. Molecular

Ecology 3:15-21.

Linder, CR. and J. Schmitt 1994. Assessing the risks of transgene escape through

time and crop-wild hybrid persistence. Molecular Ecology 3:23-30.

. 1995. Potential persistence of escaped transgenes: Performance of
transgenic, oil-modified Brassica seeds and seedlings. (In press) Ecological

Applications.

Manasse, R 1992. Ecological risks of transgenic plants: Effects of spatial dispersion

on gene flow. Ecological Applications 2(4):431-438.

. and P Kareiva 1991. Quantifying the spread of recombinant genes and
organisms. pgs. 215-231, In Ginzburg, L. [ed] Assessing Ecological Risks of

Biotechnology. Butterworth-Heinemann, Boston, MA.

McPartlan, HQ and RI. Dale 1994. An assessment of gene transfer by pollen from
field-grown transgenic potatoes to non-transgenic potatoes and related species.

Transgenic Research 3 :216-225.

12
Miller, H.I., D.W. Altman, J.H. Barton and S.L. Huttner 1995. An algorithm for the

oversight of field trials in economically developing countries. Bio/Technology

13:955-959.

Moffat, AS. 1994. Developing nations adapt biotech for own needs. Science

265:186—187.

Mooney, HA. and J.A. Drake [eds] 1986. Ecology of Biological Invasions of North

America and Hawaii. Springer-Verlag, New York.

1990. The release of genetically designed organisms in the environment:
lessons from the study of the ecology of biological invasions. In: Introduction
of Genetically Modified Organisms into the Environment [eds] Mooney, H.A.,
G. Bemardi; Scientific Committee on Problems of the Environment 44), pp.

117-127. John Wiley and Sons, Chichester.

Morris, W.F., P.M. Kareiva and PL. Raymer 1994. Do barren zones and pollen traps

reduce gene escape from transgenic crops? Ecological Applications 4(1):159-

165.

Raybould, AF. and A]. Gray 1993. Genetically modified crops and hybridization

with wild relatives: A UK perspective. Journal of Applied Ecology 302199-219

l3
Raybould, AF. and A]. Gray 1994. Will hybrids of genetically modified crops

invade natural communities? Trends in Ecology and Evolution 9(3):85-89.

Regal, R]. 1994. Scientific principles for ecologically based risk assessment of

transgenic organisms. Molecular Ecology 325-13.

Rissler, J. and M. Mellon 1993. Perils amidst the promise: Ecological risks of

transgenic crops in a global market. Union of Concerned Scientists,

Cambridge, MA.

1994. In the R&D pipeline. The Gene Exchange 5(2):7,9.

1995. Whats coming to market? The Gene Exchange 6(2&3):8-9.

Rogers, HI. and HC. Parks 1995. Transgenic plants and the environment. Journal of

Experimental Botany 46(286):467-488.

Scheffler, J.A., R. Parkinson, and RI. Dale 1993. Frequency and distance of pollen
dispersal from transgenic oilseed rape (Brassica napus). Transgenic Research

2:356-364.

l4

Scheffler, J.A. and RI Dale 1994. Opportunities for gene transfer from transgenic

oilseed rape (Brassica napus) to related species. Transgenic Research 3:263-

278.

Schmitt, J. and CR. Linder 1994. Will escaped transgenes lead to ecological release?

Molecular Ecology 3:71-74.

Small, E. 1984. Hybridization in the domesticated-weed-wild complex. in Plant

Biosystematics. [ed] W.F. Grant pgs. 195-210. Academic Press.

Tynan, J.L., M.K. Williams and A]. Conner 1990. Low frequency of pollen dispersal

from a field of transgenic potatoes. Journal of Genetics and Breeding 44:303-
306.

Umbeck, P.F., K.A. Barton, EV. Nordheim, J.C. McCarty, W.L. Parrott and IN.
Jenkins 1991. Degree of pollen dispersal by insects from a field test of

genetically engineered cotton. Journal of Economic Entomology 84(6):]943-
1950.

Williamson, M. 1994. Commrmity response to transgenic plant release: predictions

from British experience of invasive plants and feral crop plants. Molecular

Ecology 3:75-79.

15

Wright, S. 1951. The genetical structure of populations. Ann. Eugen. 15:323-354.

Zambryski, P., H. Joos, C. Genetello, J. Leemans, M. VanMontagu, J. Schell 1983. Ti
plasmid vector for the introduction of DNA into plant cells without alteration

of their normal regeneration capacity. EMBO Journal 2:2143-2150.

CHAPTER 1

CHARACTERIZATION OF THE BLUNT LEAF APEX

(bla) TRAIT IN CUCUMBER

16

17

Abstract

A further characterization of the cucumber (Cucumis sativus L.) mutant,
‘Wisconsin SMR-18 bla' (blunt leaf apex), revealed a new character associated with the
previously described leaf phenotype. The attachment of the blade to the petiole of bla
plants is flat across as opposed to the cordate or indented attachment seen in the wild type
‘Wisconsin SMR-18’ plants. The new character, truncate leaf base, was easier to score
and becomes distinctive earlier in development than previously described leaf apex
characters. It was expressed consistently in homozygous bla plants. Segregation analysis
of 1159 F2 seedlings arising from self pollinated ‘Wisconsin SMR-18'/W isconsin SMR-18
bIa' F , plants suggested that the leaf base and leaf apex characters were controlled by a
single locus or two tightly linked ones with a maximum distance between the two of 0.03
cM. In a field study of growth and fitness characteristics, the two genotypes did not
differ significantly for numbers of flowers or fruits. The similar flowering and fruiting
characteristics along with the reliable, early occurring truncate character should make the

two genotypes useful for pollination and gene movement studies.

18

Numerous genes have been described in cucumber (Cucumis sativus L.) (Pierce
and Wehner, 1987, 1990). Among these, nine are associated with distinctive leaf
morphologies (Vakalounakis, 1992). Leaf morphological mutations can be useful as
markers for hybrid production, and for pollination, genetic, and linkage studies. We have
been using the recessive bla (blunt leaf apex) mutant as a marker trait to monitor pollen
mediated gene flow from donor (wild type) to recipient (bla bla) populations (Hokanson
et al., 1994). The bid mutant was first described by Robinson (1987) as a recessive
seedling marker trait that arose from a mutagenized ‘Wisconsin SMR-l8' cucumber
population. Individuals expressing the trait were reported to have a rounded leaf apex,
rather than the pointed leaf apex typical of wild type ‘Wisconsin SMR-18'. Seeds of the
bio mutant were originally provided by R. Robinson (NY Agric. Exp. Sta, Cornell
University, Geneva, NY). Seed increases were performed by hand pollinations in the
greenhouse and by field pollinations in bee-proof cages.

In the process of working with the mutant, we found that expression of the leaf
apex trait was variable. The phenotypes observed for ‘Wisconsin SMR-l8' bla plants in
the greenhouse ranged from leaves exhibiting extremely rounded leaf apices with reduced
lobing and serration as originally described by Robinson (1987), to leaves with nearly
pointed apices, and sufficient lobing and serration to blur the distinction between bid and
wild type (Fig. 1A). The lobing and serration traits were quite variable. The wild type
‘Wisconsin SMR-18' also could exhibit variable characteristics in the first true leaf. In
a test seedling population, the apex of the first true leaf was blunt in 9 of 33 wild type
‘Wisconsin SMR-18' plants.

Here we describe a second characteristic associated with the leaf genotype, a flat

19

Figure 1. Bla Phenotypes. A. Variation in leaf shape for first true leaves of
'Wisconsin SMR-18' bla. Arrows indicate a 'Wisconsin SMR-l8' bla
individual with a first leaf with an acute apex and truncate leaf base. B.
Leaf phenotypes for 'Wisconsin SMR-18' (left) and 'Wisconsin SMR-18' bla
(right). Note the indented or cordate leaf base for the 'Wisconsin SMR-l8'
leaf on the left and the truncate base for 'Wisconsin SMR-18' bla on the
right. C. Progeny from an open-pollinated 'Wisconsin SMR-18' bla plant.
White dots indicate wild type 'Wisconsin SMR-18' leaves originating from a
heterozygous seedling (an outcrossing event). Note other leaves have the

recessive truncate leaf base.

20

 

Figure 1.

21

or truncate leaf base. The attachment of the blade to the petiole of bla plants is flat
across, rather than indented or cordate (Fig. 1A, B). This trait was readily observed in
the second and all subsequent true leaves. The truncate leaf base was consistently
expressed in homozygous bla plants (Fig. 1A). Regardless of leaf shape (i.e., apex,
serration and lobing), if the first two leaves had truncate leaf bases, the plants always had
the bla phenotype at maturity. Progeny of self-pollinated bla plants always exhibited the
mutant phenotype. Although the first true leaves of some wild type ‘Wisconsin SMR-l8'
plants (7 of 33) had a truncate leaf base, all subsequent leaves had cordate leaf bases and
pointed leaf apices. Similarly, heterozygotes clearly exhibited the dominant, cordate leaf
base (Fig. 1C).

To verify that the truncate leaf base trait was due to the presence of the bla mutation
rather than a mutation at a separate locus, we self-pollinated flowers on 11 ‘Wisconsin
SMR-18'l ‘Wisconsin SMR-18' bla F, plants. All 11 of these F, plants exhibited both
the dominant cordate leaf base and wild type leaf apex. We evaluated 1159 seedlings
arising from 16 different fruits from the 11 F, plants. Seedlings were assessed at the
first, second, and third leaf stage for leaf base and leaf apex characters (Table 1). At the
first leaf stage, approximately 10% of the wild type individuals resembled the mutant
either for the shape of the leaf apex, the leaf base, or both; the apex was more variable
than the base. When the second and third leaves were scored, both the leaf base and apex
characters segregated in the expected 3:1 ratios for a recessive, single gene trait.
Importantly, from the second leaf on, there was complete correlation between the two
traits; no recombinants (blunt leaf apex associated with cordate leaf base or acute leaf

apex with tnmcate base) were observed among the 1159 F2 seedlings. These results

22

23.5332 £96— So. 98 3. 05 «a Seneca mo oouwon one new Eco—Mama

.83 3355 no 38% :53 82835 33:29: 05 new 2.3 coaewawg mm 3.398 5 new .3. 9323-29

 

 

Scum Relax
:33 $33 ”2 c 9 am am

ageing ”annex
Seam commas com o a ma 2N

....%suflx .....m.$u~x
Secs £32. an a. .. Sn 1

080.556.; 2536:»? 3.3 2355 3.3 25 33 08955 83 23»

Dam is < 3% :55 use .53 3% 2.3 gas 33 528.. as:

 

8.3 meo— 0328 can some .32 :53 we 5.3303860 c8 :6... ._ 035.

23

suggested that the leaf base and leaf apex characters were controlled by a single locus or
two tightly linked ones (maximum distance 0.03 cM; product ratio method). Since
expression of the bla and wild type phenotypes was more variable for the first leaf than
for later leaves, it is important to score these traits no earlier than the second true leaf.

To further analyze the utility of this mutation, the bla mutant also was studied for
growth and fitness characteristics in the field using a randomized complete block design
with four replications. The two genotypes did not differ significantly for numbers of
flowers or fruits (Table 2). Comparable flowering and fruiting characteristics between the
two genotypes should allow for pollination and gene movement studies to be
accomplished without bias due to diminished reproductive performance of one of the
genotypes.

In summary, the occurrence of the truncate leaf base character increases the
usefulness of the bla mutant for screening large seedling populations, especially in the
early seedling stage. Although both characters (blunt apex and truncate base) are reliable
in later stages of development, the leaf base character becomes distinctive and consistent

sooner in development than does the leaf apex.

24

.3d A m .355.» .«o mien—Ea new 023i Enema—$6523

.33 .3 cone—390m do 05> 05 :0 38 n NV 315.3 838:: 30.?

.mBBoc 2a Bow .«0 Bean: new bung—hams 3&6 .0: Be On? mambo—5w 03. 2; A32

.2 Ba 3 d sewn}, £82. nematomno 025 B>o con—8:3 :83 Ba E033.“ 32:53 done we Conan: 38.?
.958 :00an S mg .958

E55 8 nd 38% 803 353 .83 con 3.80.» 353 35:00 cog .83 non 353 RE £83 08238 Son

 

 

a. 32 a 2 son .8 35a 3.;
a: 938.? we 3 ”Ea—m can EoBoa 232
Em: \ am: as 2.5,; as? 2.5% 23. sec.

 

sunbeam as 23.5 3288.3. 15.2-3.5 52823. E madman Ea 3:038 do sewage .N as...

25

Literature Cited

Hokanson, S.C., J.F. Hancock, and R. Grumet. 1994. Border rows as a possible
containment strategy to limit the spread of transgenic pollen. Am. J. Bot. 81:56.

Pierce, LK. and TC. Wehner. 1987. Gene list for cucumber. Cucurbit Genet. Coop. Rpt.
12291-103.

Pierce, L.K. and TC. Wehner. 1990. Review of genes and linkage groups in cucumber.
HortScience 25:605—615.

Robinson, R.W. 1987. Blunt leaf apex, a cucumber mutant induced by a chemical
mutagen. Cucurbit Genet. Coop. Rpt. 10:6.

Vakalounakis, DJ. 1992. Heart leaf, a recessive leaf shape marker in cucumber:

Linkage with disease resistance and other traits. J. Hered. 83:217-221.

CHAPTER 2

EFFECT OF BORDER ROWS AND TRAP/DONOR RATIOS

ON POLLEN-MEDIATED GENE MOVEMENT

26

27

Abstract

One frequently voiced concern associated with the field testing and agronomic
scale release of transgenic crops is the potential for pollen-mediated escape of
engineered genes into naturally occurring populations of wild relatives. Border rows
have been commonly used for restricting the pollen-mediated escape of engineered
genes in field tests. However, the efficacy of border rows for restricting such gene
movement has been little studied. Isogenic lines of cucumber (Cucumis sativus L.)
differing for a seedling marker trait, blunt leaf apex (bla) were planted in various trap
plant to donor plant ratios to test border rows as a means to control pollen-mediated
gene movement out of plantings of insect pollinated crops. All treatments had donor
plants, (Wisconsin SMR-18 with the wild type leaf shape) in the center of the plot.
Three of the four treatments had border plantings of recipients, ('SMR-18bla')
surrounding the donors. For each of the plots, groups of four recipient plants
(satellites) were planted fifty meters from the plot center in eight directions. Progeny
of the recipient plants from the satellites and borders were screened to determine the
percentage of outcrossing as measured by occurrence of the dominant phenotype.
Gene movement within the plot borders assumed a leptokurtic distribution. At each
distance from the plot center, there was more gene movement in plots with 2m2 donors
than 1m2 donors. Long distance gene movement to the satellites increased

significantly as the trap/donor ratio decreased. These results suggest that border rows

28

might serve to control the movement of transgenic pollen in small experimental plots.
However, variability in amounts of gene movement to individual satellites within
treatments (ranging from 0-3 8%) suggests environmental variables might render
predictions concerning gene flow levels and containment strategies quite difficult.
Moreover, to achieve the trap/donor ratios of the most protective treatment in these

experiments on a commercial scale would in all likelihood be economically infeasible.

Introduction

Pollen mediated escape of engineered genes into the environment frequently
has been mentioned as a potential risk associated with the large-scale release of
genetically engineered transgenic crops (Colwell et al., 1985; Ellstrand, 1988; Ellstrand
and Hoffman, 1990; Dale, 1992; Raybould and Gray, 1994; Rogers and Parks, 1995).
A commonly described scenario involves the movement of engineered genes into wild
or weedy relatives. Crop/weed hybridizations have been documented and can involve
significant amounts of long distance gene movement. For example, Klinger et al.
(1991,1992) measured levels of hybridization between cultivated and wild radish,
Raphanus sativus L. at distances up to 1,000 m. Kirkpatrick and Wilson (1988)
reported hybridizations in both directions between cultivated Cucurbita pepo and
naturally occurring Texas gourd Cucurbita (exam at distances of 1,300 m. Wilson
and Manhart (1993) found high rates of hybridization between cultivated Chenopodium
quinoa (Andean grain chenopod) and a North American wild relative C. berlandien' at
500 m from the cultivated plot.

With regard to the concerns associated with transgenic plants, several schemes

29

have been proposed to restrict pollen mediated escape of engineered genes. These
include: 1) isolating the genetically engineered crop by distance from wild relatives, 2)
surrounding the genetically engineered crop with rows of non-engineered pollen trap
plants, and 3) genetic isolation of the transgenic cr0p using mechanisms such as male
sterility, (if the fruit or seed is not the product of commerce), or linking the engineered
gene to a pollen lethal gene (Ellstrand and Hoffman, 1990; Kareiva et al., 1994). To
date, isolation by distance, and the use of border rows of pollen trap plants have been
the most widely utilized methods to satisfy USDA requirements for field testing
transgenic plants (Wrubel et al., 1992).

The use of isolation by distance and border rows to limit the spread of
transgenic pollen stems from the historical precedent of using these methods to ensure
genetic purity in seed multiplication plots (George, 1985; Kelly, 1988). While these
methods have been used successfully to prevent undesirable pollen from moving into
seed multiplication plots, the question remains as to whether they can control the
movement of genes out of plots. Several recent studies have suggested that isolation
by distance may be an unreliable method to control pollen mediated gene escape.
Manasse (1992) found that increasing isolation distances from 0.5m to 4m between
individuals or groups of Brassica campestn's also increased mean gene flow. When
Morris et al. (1994) compared the use of border rows to barren zones as a means to
control the spread of transgenes from Brassica napus, they concluded that the use of
barren zones might actually increase the amounts of gene movement over that which
would be expected if the same space were planted with a crop. They also found that

gene movement levels differed between locations, signaling the key role played by

30

environmental variables in gene movement.

Border rows have been used as a method to maintain genetic purity in wind
pollinated seed crops such as beet, Beta vulgar-is (Dark, 1971) corn, Zea mays (Kelly,
1988) and some insect pollinated crops such as cotton Gossypium sp. (Green and
Jones, 1953). Isolation results from either a physical blocking of foreign pollen by the
height and density characteristics of the border crop or by diluting the foreign pollen
with ”non-polluting" border row pollen. A simple extension of this thinking leads to
the idea that these methods might also serve to keep pollen from moving out of a
bordered plot.

The few studies testing border rows as a means to control movement of genes
out of a plot have focused on gene movement into contiguous border plantings. Tynan
et al. (1990) reported low levels of gene movement from transgenic Solanum
tuberosum into wild-type potatoes planted within the transgenic trial and in contiguous
border plantings of the wild-type. Values ranged from approximately 1% among the
inter-planted wild and engineered types to 0.05% at 3-4.5m in the border; no
movement was recorded beyond this distance. Scheffler et al. (1993) measured
movement of a transgene from a 9 m circular center plot of engineered Brassica
napus into a contiguous 70 m2 border of non transgenic B. napus. They found
negligible amounts of movement of the marker gene beyond 6 m, (less than 0.03%)
and no movement beyond 36 m. Umbeck et al. (1991) reported a "consistent and
significant" reduction in outcrossing into a 25 m wide border as the distance from the
donors increased. Outcrossing ranged from nearly 5% at the border of the donors to

less than 1% at 25 m.

31

Although these studies indicate that most pollen is distributed within a short
distance from the donors and should therefore be trapped by borders, they do not
address the question of gene movement into non-contiguous plantings which would
more closely mirror the situation for patches of wild relatives. When Handel (1982)
used a dominant morphological marker to measure gene movement within an 18 m2
plot of cultivated melons (Cucumis melo), he found that gene movement was
asymmetrical with occasional large numbers of dominant seedlings arising from fruits
at the edge of the plot. In similar experiments with cucumber (Cucumis sativus), he
found clumping of genes on the edge of 25 m2 and 16 by 12 m plots (Handel, 1983).
The aggregation of genes on plot edges may be due to an ecotonal effect or they may
suggest that pollinators carry marked pollen out of the plot.

While all the above studies have generated important information on patterns of
gene flow, no controlled experiments have been designed to test the effectiveness of
border rows in preventing long distance gene movement to discontiguous plots. Also,
no studies have directly tested the effects of varying relative donor plant to trap plant
ratios on gene movement. In this study we investigate varying donor plant to trap
plant ratios and predict that as trap/donor ratios decrease, more pollen would escape
due to an overloading of the trap plants with donor pollen. In the experiments
described herein, we used a morphologically marked cucumber genotype to address the
following questions: 1) What is the frequency and pattern of gene movement into the
contiguous border, 2) What is the effect of varying relative donor plant to trap plant
ratios on the rate of gene movement, and 3) Can border rows effectively limit long

distance movement of donor genes as measured by discontiguous satellite plots?

32

Materials and Methods

BMW. The monoecious isogenic cucumber (Cucumis sativus L.) lines
Wisconsin SMR-18 and SMR-18 bla (blunt leaf apex) were used to monitor pollen
movement. Cucumber is a predominantly outcrossing crop, 23-77% in the field,

(W ehner & Jenkins, 1985), that is pollinated primarily by honeybees, Apis melhfera
(Free, 1993). SMR-18 bla is a recessive mutation which arose from a mutagenized
Wisconsin SMR-18 population (Robinson, 1987). Individuals expressing the bla trait
have reduced lobing and serration, a rounded leaf apex rather than the pointed leaf
apex typical of wild type Wisconsin SMR-18, and a flat leaf attachment as compared
to the indented attachment of Wisconsin SMR-18. The leaf attachment character
serves as a reliable, readily scorable seedling marker trait (Hokanson et al., 1995).
Wisconsin SMR-18 and SMR—18 bla were shown to have comparable flowering and
fruiting characteristics (Hokanson et al., 1995). Wisconsin SMR-18 seed was
purchased from Agway Inc. (Syracuse, NY). SMR-18 bla seed was originally
provided by R. Robinson (New York State Agriculture Experiment Station, Cornell
University, Geneva, NY) and subsequently field multiplied in bee-proof cages.

Field experiments and greenhouse screening. The dominant, wild type
genotype Wisconsin SMR-18 was used as the pollen donor, and the recessive SMR-18
bla as a pollen trap or recipient. The overall plot design is depicted in Figure 1.
Gene movement was detected by screening for dominant type seedlings among
seedlings originating from recessive plants in the plot borders and satellites.

Four treatments varying in ratio of trap to donor plants were utilized to test the

33

Figure 1. A. Overall field plot design used in 1992 and 1993 experiments. All 12
plots regardless of design were surrounded by eight satellite plots 50 meters distant,
each containing four recessive SMR-18 bla plants. B-E. Four treatments tested in
these experiments depicted from highest to lowest trap to donor ratio. B. 1m2 of
SMR-18 donors surrounded by a 400m2 border of recessive SMR-l8 bla plants (ratio
131.1). C. 1m2 of donors surrounded by a 100m2 border of recessive recipients (ratio
34.0). D. 4m2 of donors surrounded by a 100m2 border of recessive recipients (ratio
11.6). E. 1m2 of SMR-18 donors with no surrounding border plants (borderless)(ratio
0.0). At harvest borders were subdivided into 1m2 subplots and 2-3 fruit were
collected from each subplot (see also Fig. 3). A Location of bee hives on the edge of

the donor plots.

 

Figure l.

34

 

 

 

 

 

\\\\\\\\\\\\\\\‘

}

77T
III
III
III
III
III
III
III
III
III
III
III
III
III
III
III
I

\\\\\\\\\\\\\\\

\\\\\\\\\\\

 

‘3‘\\\\‘\\‘\\\‘\“‘V‘V‘\\““\\\\

 

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
t\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\V
)\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

l

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\V
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

l

p\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\V

1\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\‘

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\V

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\‘

I

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\V

L

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\‘

)\\\\\\\\\\\\\\\\\

\\\\\\\\\\\\\\\~4

\\

\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ V
\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ N
L \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ ‘

\\\\\\\\\\\

 

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\V
p\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\V

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\‘

l

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

)\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\‘

i\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\‘

):\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\V
-

’7”7”””r777r
r IIIIIIlI/IIII
VI Ill/II/lI/III
f I/IIIIIIIIIIII
VIIIIIIIIIIIIIIII
III/IIIIIIIIII/Il
VII/IIIIIII’IIIII
III/II IIIIII’III
VIII/I I-IIIIIII
fIIIIIII III/III
VIII/IIIHIIIIIII
VIII/II I III/Ill
V’II’IIII IIIIIII
VIII/II/II/IIIIII
V’IIIIIIIIIIIIIII
Ill/ll '-/////
HIM/10mfi
rrrrrrrrrrrrrrr—V

IIIIIIIIIIIIII

Ill/IIIIl/I/I
y I’ll/IIIIIII/I
VIII/111111111111
VIII/IIIIIIIIIIII
VIIIIIIIIIIIIIIII
VII/z‘ III/II
VIII/II IIII/I
III/III III/II
VIII/II III/II
rill/IIHIIII/I
fIIII/II IIIIIII
VIII/IIzmIII/III
VIII/IIIIIIIIIIII

lil/III‘OmIIII/a

 

 

 

 

 

 

35

efficacy of border rows: 1) 1 m2 of donor plants surrounded by 400 m2 of border
plants, 2) 1 m2 donors with a 100 m2 border, 3) 4 m2 donors with a 100 m2 border,
and 4) l m2 donors with no border (Table 1). The 400 m2 border plots were
comprised of 29 rows with 41 plants per row; 9 donors were placed in the middle of
the center 3 rows. The 100 in2 border plots had 15 rows with 21 plants per row with
9 or 25 donors in the middle of the center 3 or 5 rows respectively. The borderless
plot contained 9 donor plants surrounded by a 100 m2 cultivated, unplanted area. All
of the donor populations, regardless of treatment, were surrounded by eight satellite
plots which were located 50 m from the plot center (Fig. 1). Each satellite plot, which
measured approximately 1.2 m x 1.2 m, contained four cucumber plants with the
recessive bla phenotype.

All four treatments were tested in the summer of 1992, each replicated in two
locations (Table 1). In the summer of 1993, two of these treatments were repeated at
two locations, the 100 m2 border with l m2 donors and the borderless plot with 1 m2
donors (Table l). Planting dates were June 15 and 16 in 1992, and June 16 in 1993.

All plots were isolated by at least 1500 m from any other cucumber plantings.
The plots were maintained in a manner similar to commercial plantings (Motes, 1977).
During preparative cultivation, 16°16'16 fertilizer was incorporated into the soil at a
rate of 452.5 kg per hectare. Approximately six weeks later, the rows were
sidedressed with nitrogen at a rate of 45 kg per hectare. All plots were hand
cultivated until a full canopy developed. About 1.3 cm of water per week was applied
to the plots in the absence of adequate rain. When the plants were in full flower and

female flowers were opening, (August 20 in 1992 and July 23 in 1993), one bee hive

36

.5 5550 .5ng .85 5525 saw Swag: .32.
.E .580 83.53 .833 32835 32.3— .mmm...
00.0 .8006 5 £520 «0 308050033 05 5 £520 m0 308:?

 

 

a saw: 53 55 3583.

8 "we. 23. Nag..— 30:3.

Na .82 exam $2 5383. $43 2.2.: as 328 oz

8 "men ban gm ban $23 3 5:8 Ne _

8 “BE 380 sea 883 =8 2a 85 38m .028 as 2

a Nmax 23 8.5 $23 0.: .880 ”a v

a mom: 8m 85m 322.5 sfiésfisom

8 "me. New: 2a 533% 328.com E2 was

a Maw: #82 83m sea: $2 5:3 Na 2:

Na "mmx 35.: am 538% 308.80 85 m8,— osm :88 Na _

3 sum: Sm Ban .3233 ssésfism ES: 322 "8 2:.

a "men a: cause :52 x3 :2 08% ”a z
30> 00:30; 503503“? 2% 0:8 000023; 00253;

 

.83 05 N03 5 :08 00 80:00:02 05 00.“ 30830— 000 mote. 55 3:25.30. 50m 05 .«0 :03 com 038 ScoEQufi. A 030,—.

37

containing approximately 35,000 honeybees, Apis mellifera was placed on the edge of
the donor plants with the hive opening facing southeast (Fig. l). The hives were
removed prior to fruit harvest ( September 18 and August 31 respectively).

To evaluate long distance gene movement, fruits were harvested from the
individual satellites, and stored in plastic bags at 05°C until seeds were extracted.

The plots were harvested on October 5 and 6 in 1992 and September 8 and 15 in
1993. Seeds were extracted from each fruit, air dried, bulked by satellite in paper seed
envelopes, and stored at 10°C and 25% relative humidity.

The seeds were germinated and scored in a greenhouse at Michigan State
University, East Lansing, MI, beginning in January of 1993. The seeds were planted
75 to a tray (15 seeds per row, 5 rows per tray) in 56 cm x 28 cm plastic seedling
trays. The seedlings were grown in a soil mix of 1 sphagnum peat perlite mix (Baccto
Professional Planting Mix, Michigan Peat Co., Houston, TX): 1 sterilized sandy loam.
From mid-November through mid-April, the seedlings were grown under artificial light
ranging in intensity from 258 ;.imol-s“-m2 photosynthetically active radiation (PAR) on
an overcast day to 530 nmol-s"-m2 PAR on a sunny day.

The plot borders were divided into 1 rn2 subplots to evaluate short distance
gene movement. Fruits were harvested from each subplot on the same dates as the
satellites. Fruit and seed were processed as described above.

mm. Both short and long distance gene movement was expressed as
the percentage of dominant type seedlings appearing among the seedlings germinated
from recessive parent plants in the subplots and satellites respectively. To analyze

short distance gene movement within the borders, the subplots were grouped according

38

to their distance from the plot center. To test the effect of increased donor plot size
on percent outcrossing within the border planting, paired comparisons were made at
each distance from the plot center (2 m - 7 m) for which direct comparisons could be
made. Analyses were performed on data from pairs of plots (1 m2/ 100 m2 and 4
m2/100 m2) at the same location (KBS or MSU) in the same year. The data [12 pairs
of l m2 and 4 m2 donor plots with 100 m2 borders (6 distances, 2 locations)] were
analyzed both by T-test of paired observations and Wilcoxin's signed rank test for
paired observations. Both tests gave the same results. To evaluate long distance gene
movement, mean percent gene movement into the satellites was plotted for both
seasons as a function of trap/donor ratio. Data were analyzed by Spearman's
coefficient of rank correlations and by regression analysis on arcsin linearized data as

per Steele and Torrie (1960).

Results

Shortdistance gene movement. Gene movement within the borders assumed a

 

leptokurtic distribution at all sites (Fig. 2). The highest percent gene movement,
74.0% and 91.0% were observed at the closest distances, one meter (34.0 trap/donor
ratio) and two meters (11.6 trap/donor ratio) respectively. Values decreased rapidly
with increasing distance from the plot center. Similar to what was observed by
Handel (1982, 1983), individual plot maps revealed a few instances of high levels of
outcrossing occuring at the plots' edge (Fig. 3). There were no consistent trends for
overall gene distribution within the borders that could be attributed to the location of

the hives.

39

Figure 2. Comparison of short distance gene movement (percent gene movement
observed within the borders at one meter intervals from the plot center) for plots with
lmz, A, or 4m2, 0 donor plots. One square meter data are the mean of all plots with
1m2 donors in 1992 and 1993, 4m2 data are the mean of the two 4rn"‘/100m2 plots in

1992. Data for paired comparisons came from 1m2 and 4m2 donor plots with 100m2

borders at two locations in 1992.

 

40

 

 

ax; E9532 0:06

12 14 16 18

10

Distance (m)

Figure 2.

41

Figure 3. Representative example of gene movement into the border subplots (plotted
as % wild type seedlings among all seedlings in each 1 m2 subplot) for the 1992 Kell-

1 plot. Each square represents a 1m2 subplot.

42

AmeA
$3-5
$8-: .
$2; ,.
o\oo
@2953 82

$5803 25 2:.» .x.

 

Figure 3.

43

Movement from the l m:2 and 4 m2 donor plots (34.0 and 11.6 trap/donor ratios
respectively) had similar distribution patterns (Fig. 2). However, there was
significantly more gene movement into the plot borders from the 4 m2 donor plots than
the 1 m2 donor plots (p5 0.005; T-test or Wilcoxin sign rank test of 4 m2 and l m2
donor plot data paired by distance from the plot center).

Lgngglisjance gene movement. As was the case with short distance gene
movement, there was no apparent association between hive placement and patterns of
long distance gene dispersal to the satellite plots. Gene flow to the satellites was
highest in the absence of border/trap plantings (Fig. 4; Table 2). In general, long
distance gene movement into the satellites decreased significantly as the trap/donor
ratio increased (Table 2; Fig. 4, linearized regression y = 8.11 - 5.23x, df = 10; r2:
0.671, ps 0.05 ). The Spearman's coefficient of rank correlation was r, = 0.906. The
percent gene movement into the satellites for the plots with a trap /donor ratio of 34.0
ranged from 0-0.18, the 11.6 ratio plots ranged from 0.38-0.83, while the borderless
plots ranged from 0.68-4.7. Gene movement was not detected for the 131.1 ratio
plots.

Gene flow into the individual satellites was generally low (Table 3). All of the
plots with trap/donor ratios of 131.1 or 34.0 had no satellites with gene movement
over 1%. Among those plots with trap/donor ratios of 11.6 and 0, values generally
ranged between 0 and 4.7 percent. Overall, the percent gene movement was evenly

distributed among the eight satellites surrounding each donor plot (Table 3).

44

Figure 4. Mean values (:t standard errors) for long distance gene movement (% of
total seedlings with the dominant marker trait) to satellite plots for the four treatments

tested in 1992 and 1993, plotted as a function of the trap/donor ratio for the treatment.

 

3.0

2.5

2.0

1.5

1.0

Gene Movement (%)

0.5

0.0

Figure 4.

 

45

 

 

I ' I ' l ' l ' I ' l f l T l

0 20 40 60 80 100 120 140

Ratio of Traps/Donors

46

3:32 sumo 8 8:28am 2mm. Soc was?» :32.

 

 

o; :2 8 53

NE «a: 8 23M

o: 84 Na mecca

woo cm: .8 55 ed 522. 2. Ne _

So as. No 820 sesame o2

woo E 8 7:3. 9: 8:8 ”a e

8e 32 8 5mm:

36 $8 8 8m

23 a: 8 fig: 522 "a 2:

m3 3: 8 SEE 3;. 8:8 ”a _

23 in 8 8m 328 NE 2:

85 >8 8 x3 22 SEES _
$58052 280 ..\o cocoouom n 30> 25 .8533; 308:3;

 

.33 can 83 E 35:58.: 38 05 8m 32a 2538 05 9 2.08968 ocow 3.33% 9.2 532 .N 033.

47

0.008 030$ 02—

 

 

00.0 00.0 00.0 00.0 02 8.0 0: 8.0 9.0 0.0 82 an.”
0: 03 8.0 03 00.0 0: 00.0 E 8.0 0.0 82 2.3.
0:. 00.0 00.0 00.0 00.0 0.8 00.0 00.0 00.0 0.0 8.: 8000
00.0 0:. 2.0 00.0 00.0 00.0 00.0 00.0 02 0.0 82 ES
80 8.0 00.0 00.0 00.0 8.0 8.0 00.0 0:. 0.0 _ N8— 806
8.0 00.0 00.0 00.0 0: 00.0 3 00.0 8.0 0.: 8.: _.=3_
00.0 00.0 00.0 00.0 00.0 00.0 8.0 00.0 00.0 0.3 82 355
00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 :0 0.8 80. an
00.0 00.0 00.0 3 3 00.0 00.0 3 00.0 0.3 82 0.82
£0 80 00.0 :0 .00 00.0 00.0 00.0 00.0 0.3 82 7me
00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 :2 N8. .8
8.0 00.0 00.0 00.0 00.0 00.0 00.0 3 .3 :2 N02 55
532 w b o m v m N _ 0:8 5:03am; 30> 5:80..—

 

.33 93 N93 .00 088850 05 E .0202 083 02025 05 .«o :03 553» 005.33. 3.63.65
£30 05 2 35.28 :03 E awn—=38 cab 25 0\0 «5 ma commofixov 288268 28m 3:303 maoq .m 030,—.

48

The only exception was the borderless Ponds plot with a mean value for long distance
gene movement to the satellites of 4.7%. However, only one of the eight Ponds

satellites actually received any gene flow, nearly 38%.

Discussion

Our results suggest that in small research plots, border rows might serve to
effectively reduce the pollen mediated spread of engineered genes. We found that
increasing the trap/donor ratios within the plots significantly reduced the long distance
movement of genes into satellite plots. We did not detect any pollen movement
outside the borders with a trap recipient to donor ratio of 131.1, and even the least
protective border treatment, with a trap/donor ratio of 11.6, allowed an average of only
0.75% long distance gene movement.

While such borders can probably be used to minimize gene flow within small
research plots, it would be difficult to limit gene flow out of agronomic-scale
plantings. Our data indicate that borders could be "swamped" unless they were larger
than the transgenic fields themselves. We found that increasing the number of donors
within the 100 m2 border, i.e. lowering the trap/donor ratio, had two effects. First, the
amount of gene movement into the border (short distance movement), was greater at
all distances from the plot center for the 4 m2 as compared to the l m2 donor plots.
Secondly, gene movement to the satellite plots, (long distance movement), increased as
the trap/donor ratio decreased. At our most effective trap/recipient ratio, over 100
acres of non-transgenic trap plants would need to be planted for each acre of

transgenic plants to prevent gene escape.

49

Environmental variation also makes it difficult to develop strategies for
containing transgenes in agronomic settings. For example, we found that long distance
movement to individual satellites within the plots was generally low and fairly
uniform, (04.7%), however one satellite within the Ponds borderless plot received
38%. This satellite had several small lakes (ponds) located approximately 0.25 miles
beyond it. Perhaps these lakes, being the primary water source in the area structured
bee movements creating a directionality to their food foraging leaving open the
possibility that a single environmental variation may have resulted in a major shift in
gene movement.

When large numbers of transgenic crop plants are deployed, the sheer volume
of crop genes moving into surrounding wild populations could ensure their persistence
in hybrids. One of the axioms of invasion biology is that invaders are more likely to
succeed when they have a large founding population (Mooney and Drake, 1990). The
Ponds satellite with 38% gene flow is a good example of what has been referred to as
a ”low frequency, large magnitude event".

It is often argued that some of the genes proposed for incorporation into crop
genomes via biotechnology, such as those conferring tolerance to salinity, drought or
cold could persist in crop/wild hybrids by increasing the selective advantage of the
plants in which they reside. However, selective advantage or increased fitness aside,
in those instances where large numbers of crop/wild seeds are produced through a
flooding of domesticated pollen into wild populations, a few hybrids might survive
(Glidden, 1994). As previously discussed, several successful crop/weed hybridizations

have been documented in situations where the crop and its weedy relative co-occur;

50

(Daucus carota L., Wijnheijmer et al., 1989; Oryza sativa, Langevin et al., 1990; Zea
mays, Doebley, 1990; Beta vulgaris L., Santoni and Berville, 1992; Boudry et al.,
1993; Setan'a italica, Till-Bottrand et al., 1992). Moreover, when Klinger and
Ellstrand (1994) created crop/weed hybrids in Raphanus sativus L., they found a 15%
increase in fruit and seed production as compared to wild siblings, with no reduction
in the other fitness characters measured; time to first flower, early fruit production,
average seeds per fruit, and average seed mass. Thus, in at least some instances,
crop/weed hybrids can actually have a higher level of fitness than their wild siblings.
In conclusion, while the use of border rows can reduce the extent of gene
movement out of a small test plot, use of border rows as a containment strategy in
agronomic-scale plantings is of dubious value. The numbers of non-transgenics that
must be planted to significantly reduce escape is not agronomically feasible.
Moreover, even in the most protective schemes, environmental variation can result in
substantial levels of isolated gene flow. In making decisions about the large scale
release of transgenic crops, it may be more prudent to consider the nature of the gene
itself and the crop in which it is to be deployed, rather than whether or not it can be

contained.

51

Literature Cited

Boudry, P., M. Merchen, P. Saumitou-Laprade, Ph. Vernet, H. VanDijk. 1993. The
origin and evolution of weed beets: consequences for the breeding and release

of herbicide-resistant transgenic sugar beets. Theoretical and Applied Genetics

87:471-478.

Colwell, R. K., E. A. Norse, D. Pimentel, R. E. Sharples, and D. Simberloff. 1985.

Genetic engineering in agriculture. Science 229:111-112.

Dale, P. J. 1992. Spread of engineered genes to wild relatives. Plant Physiology

100:13-15.

Dark, S. 0. S. 1971. Experiments on the cross-pollination of sugar beet in the field.

Journal of the National Institute of Agricultural Botany 12:242-266.

Doebley, J. 1990. Molecular evidence for gene flow among Zea species. BioScience

40:443-448.

Ellstrand, N. C. 1988. Pollen as a vehicle for the escape of engineered genes? pgs.
530-532 in J. Hodgson and AM. Sugden [eds] Planned Release of Genetically
Engineered Organisms. Trends in Biotechnology/Trends in Ecology and

Evolution Special Publication. Elsevier Publications, Cambridge, UK.

52

Ellstrand, NC. and C. A. Hoffman. 1990. Hybridization as an avenue of escape for

engineered genes. BioScience 40:438-442.

Free, J. B. 1993. Insect Pollination of CrOps. 239! Edition. Academic Press.

George, R. A. T. 1985. Vegetable Seed Production. Longman, NY.

Gliddon, C. 1994. The impact of hybrids between genetically modified crop plants

and their related species: biological models and theoretical perspectives.

Molecular Ecology 3241-44.

Green, J. M. and M. D. Jones. 1953. Isolation of cotton for seed increase. Agronomy

Journal 45:366-368.

Handel, S. N. 1982. Dynamics of gene flow in an experimental population of

Cucumis melo (Cucurbitaceae). American Journal of Botany 69(10):]538-1546.

. 1983a Contrasting gene flow patterns and genetic subdivision in adjacent

populations of Cucumis sativus (Cucurbitaceae). Evolution 37(4):760-771.

Hokanson, S. C., J. F. Hancock, and R. Grumet. 1995. Characterization of the blunt

leaf apex (bla) trait in cucumber. HortScience 30(7):1461-1462.

S3
Kareiva, P., W.F. Morris and CM. Jacobi. 1994. Studying and managing the risk of

cross-fertilization between transgenic crops and wild relatives. Molecular

Ecology 3:15-21.

Kelly, A. F. 1988. Seed Production of Agricultural Crops. Longman Scientific, U.K.

Kirkpatrick, K. J. and H. D. Wilson. 1988. Interspecific gene flow in Cucurbita: C.

texana vs. C. pepo. American Journal of Botany 75:519-527.

Klinger, T., D. R. Elam, and N. C. Ellstrand. 1991. Radish as a model system for the

study of engineered gene escape rates via crop-weed mating. Conservation

Biology 5(4):531-535.

P. E. Arriola, and N. C. Ellstrand. 1992. Crop-weed hybridization in

 

radish (Raphanus sativus): Effects of distance and population size. American

Journal of Botany 79(12):1431-1435.

Klinger, T., and N. C. Ellstrand. 1994. Engineered genes in wild populations.
Fitness of crop-weed hybrids of Raphanus sativus. Ecological Applications

4(1):117420.

54
Langevin, S. A., K. Clay, and J. B. Grace. 1990. The incidence and effects of

hybridization between cultivated rice and its related weed red rice (Oryza sativa

L.). Evolution 44: 1000-1008

Manasse, R. 1992. Ecological risks of transgenic plants: Effects of spatial dispersion

on gene flow. Ecological Applications 2(4):431-438.

Mooney, H. A. and J. A. Drake. 1990. The release of genetically designed organisms
in the environment: Lessons from the study of the ecology of biological
invasions. in Mooney, HA. and G. Bemardi [eds] Introduction of Genetically
Modified Organisms into the Environment. SCOPE 44. pgs. 117-127. John

Wiley & Sons, Chichester.

Morris, W. F., P. M. Kareiva, and P. L. Raymer. 1994. Do barren zones and pollen
traps reduce gene escape from transgenic crops? Ecological Applications

4(1):157-165.

Motes, J. E. 1977. Pickling cucumbers, production-harvesting. Michigan State

University Extension Bulletin E-837.

Raybould, A. F., and A. J. Gray. 1994. Will hybrids of genetically modified crops

invade natural communities? Trends in Ecology and Evolution. 9(3):85-89.

55

Robinson, R. W. 1987. Blunt leaf apex, a cucumber mutant induced by a chemical

mutagen. Cucurbit Genetics Cooperative Report 10:6.

Rogers, H.J. and HG Parks. 1995. Transgenic plants and the environment. Journal

of Experimental Botany 46(286):467-488.

Santoni, S., and A. Berville'. 1992. Evidence for gene exchange between sugar beet
Beta vulgaris L. and wild beets: Consequences for transgenic sugar beets. Plant

Molecular Biology 20:578-580.

Scheffler, J. A., R. Parkinson, and P. J. Dale. 1993. Frequency and distance of pollen
dispersal from transgenic oilseed rape (Brassica napus). Transgenic Research

2:356-364.

Steele, R R and F. J. Torrie. 1960. Principles and Procedures of Statistics. McGraw

Hill, New York, NY.

Till-Bottraud, I., X. Reboud, P. Brabant, M. Lefranc, B. Rherissi, F. Vedel, and H.
Darmency, 1992. Outcrossing and hybridization in wild and cultivated foxtail

millets: consequences for the release of transgenic crops. Theoretical and

Applied Genetics 83:940-946.

56

Tynan, J. L., M. K. Williams, and A. J. Conner. 1990. Low frequency of pollen
dispersal from a field trial of transgenic potatoes. Journal of Genetics and

Breeding 44:303 ~306.

Umbeck, P. F., K. A. Barton, E. V. Nordheim, J. C. McCarty, W. L. Parrott, and J. N.
Jenkins. 1991. Degree of pollen dispersal by insects from a field test of

genetically engineered cotton. Journal of Economic Entomology 84(6):]943-

I950.

Wehner, T. C. and S. F. Jenkins. 1985. Rate of natural outcrossing in monoecious

cucumbers. HortScience 20(2):211-213.

Wijnheijmer, E. H. M., W. A. Brandenburg, and S. J. Ter Borg. 1989. Interactions
between wild and cultivated carrots (Daucus carota L.) in the Netherlands.

Euphytica 40:147-154.

Wilson, H., and J. Manhart. 1993. Crop/weed gene flow: Chenopodium quinoa

Willd. and C. berlandieri Moq. Theoretical and Applied Genetics 86:642-648.

Wrubel, R. P., S. Krimsky, and R. E. Wetzler. 1992. Field testing transgenic plants.

BioScience 42(4):280-289.

CHAPTER 3

COMPARISON OF DISPERSAL PATTERNS

OF A TRANSGENE AND A NATIVE MARKER GENE

ORIGINATING FROM THE SAME DONOR

57

58

Abstract

Despite full commercial approval of twelve transgenic crops in the US in
1995, concern is still being expressed regarding potential risks associated with
transgenic crops. A commonly voiced concern relates to the pollen-mediated escape
of engineered genes into populations of crop wild relatives. To address this concern
the scientific community has turned to a rich body of literature on pollen dispersal that
has been generated on non-transgenic organisms. However, the utilization of this
information requires the assumption that the pollen dispersal patterns of native and
transgenes will be the same. To test the validity of this assumption, we evaluated
levels of pollen-mediated gene movement from melon plants (Cucumis melo)
expressing dominant native and engineered marker genes into a surrounding
contiguous border and into discontiguous satellite plots of recessive non-transgenic
melon plants. Long distance dispersal patterns of the native and transgene in these
experiments was identical. At no time did we observe movement of one of the genes
without the other. Long distance gene movement from plots with contiguous borders
was significantly reduced relative to borderless control plots. Dispersal patterns of the
two genes into the plot borders was nearly identical. Of the nearly 4600 seedlings
screened for both morphological (presence of green cotyledons) and transgene
movement (presence of NPT 11 protein by ELISA), in no case was the NPT gene

observed in the absence of dominant Vir trait. However, 39 seedlings were green but

59
did not express NPT II as measured by ELISA. PCR analysis revealed 27 of these 39

NPT II ELISA‘ plants to contain the NPT II gene implicating transgene inactivation as
the cause of the NPT II ELISA‘ seedlings. Segregation data suggested that the
remaining 12 NPT II ELISA’ plants were most likely the progeny of a heterozygous

green, non-transgenic mother inadvertantly planted in the border.

Introduction

The commercialization of transgenic crops is now a reality. Twelve transgenic
crop/gene combinations have now been given full commercial approval and more are
soon to follow (Pmrington and Bergelson, 1995; Rissler and Mellon, 1995). Despite
the approval granted for these transgenic crops, concern is still being expressed
regarding the potential risks associated with the commercialization of genetically
engineered craps.

One of the most commonly raised concerns relates to the potential for
engineered genes to move via pollen into populations of crop wild relatives (Colwell et
al., 1985; Ellstrand, 1988; Ellstrand and Hoffman, 1990; Dale, 1992; Raybould and
Gray, 1994). Although ”escapes” from experimental-sized plantings can be minimized
by any of a number of mechanisims including: isolation by distance, enclosure by
border rows of pollen trap plants, and/or the use of genetic isolation methods
(Ellstrand, 1988; Ellstrand and Hoffman, 1990; Kareiva et al., 1994), some level of
gene escape is virtually inevitable in commercial scale plantings (Kareiva et al., 1994;
Manasse, 1992; Hokanson et al., 1996).

Given the near certainty that transgenes will escape, concern now centers on

60

the establishment and spread of transgenes in natural populations. Direct
measurements of actual transgene establishment and spread is quite limited (Crawley et
al., 1993; Bergelson, 1994; McPartlan and Dale, 1994; Scheffler and Dale, 1993;
Linder and Schmitt, 1994, 1995). This is due to the comparatively short time that
transgenes have been available and to regulations limiting the scale and duration of
field experiments utilizing them. As a result, much of the body of evidence amassed
to respond to risk assessment issues has been based on a rich body of theoretical and
empirical research on gene dispersal completed with non-transgenic organisms
(Andow, 1994; Crawley, 1987, 1990; Darmency, 1994; Gliddon, 1994; Manasse and
Kareiva, 1991; Mooney and Drake, 1990; Williamson, 1994).

An unspoken assumption underlies the use of information gained from non-
transgenic organisms in making predictions concerning the behavior of transgenes.
The assumption is that transgenes will move, in a manner analogous to that of native
genes, i.e., in accordance with the rules of transmission genetics as we understand
them today. Once a gene becomes integrated into the chromosome, regardless of the
means, it should be transmitted between individuals in the same fashion as any other
gene. While this assumption is probably valid, it remains untested.

Evidence is now accumulating that transgenes in plants are not always
expressed in a predictable fashion (Finnegan and McElroy, 1994 and references
therin). These abnormalities, known collectively as transgene inactivation, have
produced some unexpected phenotypes. In fact, the various phenotypic effects
associated with these inactivation events at times appear to result in abberant

transmission patterns. Given some of the unique sources and the sometimes

61

unpredictable expression patterns of the genes, it is not so surprising that there are
concerns that transgenes could segregate and transmit in an abberant manner (Rissler
and Mellon, 1993). Since no direct tests comparing the movement of native and
transgenes has been reported, no contention to the contrary can be offered.

In these experiments we address the issue of whether native and transgenes
have the same pollen-mediated gene dispersal patterns. We compared the levels of
gene movement from melon plants, C ucumis melo, expressing both a dominant
morphological marker gene and a transgene, into a surrounding border of non-
transgenic melons which did not contain the marker genes. Additionally, we extended
an ongoing series of experiments (Hokanson et al., 1996) that test the efficacy of

border rows as a means to control long distance gene movement.

Methods and Materials

mm Two lines of melon (Cucumis melo L.) were used in this study.
Transgenic plants expressing the NPT II (neomycin phosphotransferase) gene and
carrying a dominant morphological marker for green cotyledons or non-virescence (V)
were utilized as donor plants in these experiments. These donor plants (line ZYCP
30) were R2 progeny of previously described transgenic plants (Fang and Grumet,
1993). Non-transgenic plants homozygous for the recessive virescent (v) trait were
used as recipients. The monoecious virescent mutant (C879-J2), provided by Dr. Perry
Nugent (U.S. Vegetable Laboratory, USDA, Charleston, SC), has been described
previously (Nugent, 1987). Pleiotropic effects of the mutation include yellow

cotyledons which turn green in approximately one week and cream colored flowers

62

that fade to white. The virescent mutant has been reported to exhibit 49% outcrossing
in the field (Nugent, 1987). The two genes were tested for independent segregation
(Table 1). Progeny segregating from a testcross of a heterozygous green, NPT 11
positive plant x two virescent NPT 11 negative plants fit a 121:1:1 segregation model.

MExperiJnents The overall field plot design used in these experiments is
depicted in Figure 1A. All donor (dominant, non-virescent, transgenic plants
expressing the NPT H gene) plots were surrounded by eight recipient (recessive, non-
transgenic, virescent plants) satellite plots located 50m distant. Each satellite plot,
which measured approximately 1.2m x 1.2m, contained four recipient plants. Two
treatments were used in these experiments (Fig.1). One treatment consisted of 1m2 of
donor plants surrounded by 100m2 of recipient plants (Fig. 1B). This plot had 11 rows
with 17 plants per row. Nine donor plants were placed in the center of the middle
three rows. The other treatment had 1m2 or nine donor plants encircled by a 100m2
barren zone (Fig. 1C). Both treatments were replicated at two locations in the
summers of 1993 and 1994 (Table 2). The plots were established on June 23, 24 and
25 in 1993 and June 9, 10 and 14 in 1994. The data presented are from the 1994
season. Although similar trends were seen in the 1993 data, the summer of 1993 was
a poor growing year for melons resulting in low amounts of poor quality seed.

In order to prevent pollen contamination from other melon plants, all plots were
isolated from other melon plantings by at least 1500m. Plots were maintained in a
manner approaching commerical practices. Prior to planting, 12.12.12 fertilizer was
incorporated into the soil at a rate of 453.5 kilograms per hectare. Roughly six weeks

later, plants were sidedressed with 20.5 kilograms per hectare of nitrogen. Weeds

63

Table 1. Test for independent segregation of virescence (vir) and neomycin
phosphotransferase (NPT II)

 

 

Classes Observed Expected x2
Yellow 45 45.5 0.196"3
NPT II ELISA'

Yellow 45 45.5

NPT II ELISA+

Green 44 45.5

NPT II ELISA'

Green 48 45.5

NPT II ELISA+

 

n’Not significantly different from the expected 1:1:l:1 ratio.

64

Figure 1. A. Overall field plot design. Each main plot (center square), was surrounded
by eight satellite plots (shaded), located 50 meters from the plot center. Each satellite
plot contained four non-transgenic recessive virescent melon plants. B and C.
Expanded view of center plots. B. The bordered plot contained 1m2 of transgenic non-
virescent donor plants (small center square), surrounded by a 100m2 border of non-
transgenic recessive virescent melon plants. C. Borderless center plot contained 1m2

of transgenic non-virescent donor plants. A Location of bee hive on the edge of the

donor plots.

65

50m

 

 

 

 

Em

C

 

 

pli’l/Ifrfrllliil

////////////////
////////////////
////////////////
II/x/xxxxx/xx/l/
////////////////
////////////////
//////////////.
/////// m//.m
///////-H////.o
//////////M.I///1
//////‘.////////
////////////////
////////////////
////////////////
B/l/l/I/I/I/I/
/////////////
V/VV/V/VI/V/lr/

 

 

Figure l.

66

H2 5:on acme—8H .mgm b38383 28m naming:
H2 .3580 Seam:— dgm 00882.5 38m 53:32“
83 Snow 5 3:03 we 038350093 05 5 00:03 me 038:?

 

 

8—04— «gm ESE—0m
€00
Naom motomem flowed ed 833 o: cocoa NE _
78m macaw same—050m wee—Emesom
.8220
team "gm 23:00:83.3 £533 «.3 BEBNE 03 xenon «8 _
cote $28? 9856on05 0:8 853%; 05888.3.

 

.33 E 0208 3:253: ofi .«e :80 8.0 03:32 05 838 383309 .N 033.

67

were controlled by hand until the canopy closed. In the absence of adequete rain,
plots were irrigated at a rate of approximately 1.3cm of water per week. In 1993,
plants were set directly into the ground. In 1994, plants were planted through plastic
mulch laid out in the rows. The mulch was used to decrease moisture evaporation
from the soil and raise the root zone soil temperature. When the female flowers were
beginning to open, (August 11 in 1993; July 26 in 1994), one bee hive containing
approximately 35,000 honeybees, (Apis mellifera) was placed at the edge of the donor
plants in each plot (Fig. 1). The hives were removed just prior to fruit harvest.

Fruits were harvested on September 28 and 29 in 1993. In 1994, fruits were
harvested as they ripened beginning on September 22 and extending through October
12. Fruits were stored in plastic bags at 0-5°C until the seeds were extracted.
Following extraction, seeds were air dried for a minimum of one day and then stored
in paper envelopes at 10°C and 25% relative humidity.

To evaluate long distance gene movement, seeds were bulked by the satellite
from which they originated. In order to compare the movement of native and
transgenes, the plot borders were divided into 1m2 subplots. Fruits were harvested and
sorted according to the subplot from which they originated.

Seeds were germinated and scored in a greenhouse at Michigan State
University, East Lansing, MI beginning in June of 1994 and continuing thru October
of 1995. The seeds were planted 50 to a tray in 56 cm x 28 cm plastic seedling trays
filled with a 1:1 mix of sphagnum peat perlite mix (Baccto Professional Planting Mix,
Michigan Peat Co., Houston, TX) and sterilized sandy loam. From mid-November

through mid-April, the seedlings were grown under artificial light ranging in intensity

68

from 258 r.rmol-s"vm2 photosynthetically active radiation (PAR) on an overcast day to
530 umol-s"-m2 PAR on a sunny day. When necessary to maintain the plants longer
for further analyses, seedlings were transplanted to six inch clay pots.

Progeny Screening Progeny from the recipient plants were screened for native
and transgene movement. Native gene movement was scored as the percentage of
green cotyledon seedlings arising from non-transgenic virescent recipient parents. If
the cotyledon score was ambiguous, a second morphological trait, flower color, was
scored.

Transgene movement was detected by screening all seedlings for NPT 11
protein using a double-antibody sandwich enzyme linked immunosorbant assay
(ELISA). The NPT II assay kit was purchased from 5 Prime - 3 Prime Inc. (Boulder
CO). Assays were performed according to the manufacturers specifications. Seedlings
were sampled at the first or second true leaf stage. Leaf samples were taken by
punching discs out of newly expanded leaves with a paper punch. Discs were placed
in Corning 96 well disposable culture plates (Corning Glass Works, Corning, NY).
The plates were placed in zip-lock bags and frozen at -80°C until the ELISA tests
were completed. Replicate samples of each leaf were collected at each sampling.

The morphological and ELISA scores were compared for each individual
seedling. In the event of an apparent discrepancy between the two scores (green
cotyledon, NPT II ELISA' or vir, NPT II ELISA+) the following tests were performed
as necessary: (1) ELISA's were rerun with both the replicate and fresh leaf samples,
(2) the second morphological trait was scored, and/or (3) genomic DNA was analyzed

for the presence of the NPT II gene using (PCR) primers specific to the NPT II gene.

69
DNA Extraction .a__n_gl_PCR Analysis Genomic DNA was extracted from plant

leaf tissue using the miniprep procedure described by Dellaporta et al. (1985) with the
following modifications: 1) Quantities were reduced to be appropriate for
approximately 0.5 grams of newly expanded leaf tissue and 2) Initial extraction was
performed with a mechanical pasta roller (Atlas pasta machine, Vitantonio
Manufacturing Co., Italy). The leaves along with 850 microliters of DNA extraction
buffer were placed in a two ounce plastic, puncture proof bag (Whirl-Paks, Baxter
Diagnostics Inc, McGaw Park, IL) and run between the rollers set at the narrowest
setting. Immediately after crushing, 750 microliters of the crude extract was pipetted
from the bag and placed in an eppendorf tube on ice to which was added 50
microliters of 20% SDS. The tubes were placed in a 65°C water bath for 10 minutes
after which 250 microliters of 5M potassium acetate was added to each tube. The
tubes were placed on ice for five minutes and then spun in a rnicrofuge at 13,750 rpms
for 10 minutes. Eight hundred microliters of the supernatant was pipetted into a new
eppendorf tube to which 560 microliters of cold isopropanol was added. Tubes were
placed in a -80°C freezer for five minutes and then were spun in a microfuge for 10
minutes at 13,750 rpm. Isopropanol was poured off the samples and the resulting
pellet was allowed to dry in the eppendorf tube. After drying, the samples were
resuspended in 50 microliters of 1X TE, to which was added one-tenth volume of 3M
sodium acetate, pH 5.2. To further purify the samples, they were digested with
RNAse and extracted once with phenolzchloroformzisoamyl alcohol (25:24:1) and once
with chloroformzisoamyl alcohol (24:1). The samples were then precipitated with two

volumes of cold ethanol followed by a rinse with 70% cold ethanol. Samples were

7O

dried and resuspended in double distilled water. Resultant DNA samples generally
ranged in amounts from 4 to 20 ng/microliter.

To verify the presence of the NPT II gene, genomic DNA samples were used
as a template for PCR. Reactions were done according to the manufacturers
specifications (Perkin Elmer, Norwalk, CT). Each reaction contained approximately
100ng of template DNA and two 18 base pair primers specific to an internal 700 base
pair region of the neomycin phosphotransferase II (NPT II) gene (Beck et al., 1982).
Thermocycler conditions were 94°C melt for 1 minute, followed by 1 minute at 62°C
for primer annealing, followed by a 2 minute extension at 72°C for 2 minutes. This
cycle was repeated 45 times, followed by a 5 minute extension at 72°C. Each set of
DNA extractions and PCR reactions included positive (NPT ELISA+ plants) and
negative (virescent NPT II ELISA’ plants) controls. In several instances DNA from
PCR+ and/or PCR‘ plants was reextracted and the samples were retested to verify
reproducability of the results.

Data Analysis Native gene movement was calculated as the percentage of
green cotyledon or yellow flowered seedlings arising from the progeny of yellow
cotyledon, white flowered parents. Similarly, transgene movement was scored as the
number of NPT II positive seedlings found among the total seedlings from non-
transgenic parents. To evaluate the influence of border rows on long distance gene
movement, we compared levels of gene movement to the satellite plots for the
borderless and bordered plots using a one—tailed t-test for samples with unequal

variance; each of the individual satellites was used as a replicate.

71

Results and Discussion

Long distance gene movement from the bordered plots to the satellites was
identical for both the native and transgene (Table 3). In each case where there was a
non-virescent seedling, there was detectable NPT 11 protein; for each virescent
seedling, no NPT II protein was detected. For both the morphological marker gene
and the transgene there was significantly less long distance gene movement to the
satellite plots from the bordered plots (0.0% - 0.11%) than from the borderless plots
(0.27% - 2.16%)(Table 3). These results are in good agreement with previous studies
on cucumber where borders were found to reduce but not eliminate gene flow from
small plots (Hokanson et al., 1996).

Short distance movement of both the native and transgene within the plot
borders assumed a leptokurtic distribution (Fig. 2). Movement of the transgene
mirrored the pattern for the native gene. As was the case for long distance movement
to the satellite plots, movement of the transgene was never detected without
concommittant movement of the native gene, i.e. there were no virescent seedlings
expressing the NPT II gene (Table 4).

There were, however, 39 green seedlings arising from a few subplots at the
Sandhill farm that did not express the NPT II protein as measured by ELISA (Fig. 2,3;
Table 4). Possibile explanations for the presence of the NPT II ELISA' green
seedlings are: 1) The gene was present, but not detectable by ELISA, i.e. there was
transgene inactivation, or 2) The gene was absent due to one of the following reasons:
a) There was a donor plant heterozygous for the NPT II gene, b) A non-transgenic

green plant could have been accidently planted in the Sandhill border plot, or c) The

72

.5528 $8830 :0 8385 $59.55 0353 E .52 8 5235 58m 0:00.23 05 8 550m.
Sum—0 .358? 3:55 5.05 538.8 no.0 .53 5:8 28 .modvq...

 

 

0\0£N $3N «mm 8—0.3

o\0h~.o ova found m3. NLom 533 oZ

$2.0 $2.0 So 3-6m

0\0o.o .mod .xbd 8: 05m 588 «8 2:
055.503. 522 2,52 558% 53832 85 «58050.3.

 

n825533 050 0\0

 

.53
8 55. Bow 2: .8 0.83 02:23. 05 8 059.58 v5 35: 05 3m 058058 358% mead .m 25,—.

73

Figure 2. Percent gene movement for the native (Vir), O , and the transgene (NPT II),
A, plotted as a function of distance in meters from the donor plot center. Gene
movement is calculated as the percent Vir (native) and percent NPT II (trans) postive

among all seedlings scored.

74

 

 

100 —

80-

?E 05.555. 850

___
050
1

Distance (m)

Figure 2.

75

Table 4. Phenotypic combinations for progeny of recipient, non-transgenic
plants. Presence of NPT II was analyzed by ELISA.

 

 

 

Phenotype

Native gene Transgene # of individuals

Vir NPT II ELISA' 3,860
(recipient phenotype)

Vir NPT II ELISAJ 0

green NPT II ELISA' 39‘

green NPT II ELISA+ 586
(donor phenotype)

 

lAll occurred in the Sandhill plot. See Figure 3.

76

Figure 3. Plot map depicting the Sandhill border subplots (numbered 1—132), and the
location of the fruits which produced the 39 NPT I] ELISA‘ green seedlings (shaded
subplots). The six striped subplots in the center depict the location of the donor
plants. Numbers in the shaded subplots represent; (Top), total number of green NPT
II ELISA“ seedlings originating from the subplot, (Bottom), total number of NPT II

ELISA' seedlings found to contain an NPT II gene with PCR.

77

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I11 111 . ”a 9' a 7? 44 AW 5 IL.—
A/ n at g,— I 13.— ‘
.4!!— {.1 4L ' r.—
L l M # n I
L M I! 1

1‘} Jo I I7 6

I“ Mr 4" 1) at J! I I!- ,g
E H4”, ,4 4,1 11 45 J? 'l
72: Id 0 " J? a} 4T": «I w r

M lol- It! 90 11 69 6 .vl a.

”I no “9 13 ‘7 I6 ”.5 W H! u. 1

 

78

native marker gene was moving in the absence of the transgene and violating normal
rules of transmission genetics.

In order to determine whether transgene inactivation was responsible for the
occurrence of the apparently NPT H ELISA' seedlings, genomic DNA was extracted
and used as a template for amplification of the NPT II gene via PCR. For 27 of the
39 NPT 11’ plants PCR analysis revealed the presence of a band corresponding to the
predicted 700 base pair NPT II fragment (Fig. 4). The presence of the NPT II gene
fragment in the absence of detectable NPT 11 protein indicates transgene inactivation
as the explanation for the apparent discrepancy between the native and engineered
gene. These results suggest that transgene inactivation occurred in at least one donor
plant. We have also observed other transgenic melon lines for which NPT II gene
expression has been silenced (Grumet, unpublished).

To further verify that the 12 remaining NPT II ELISA‘ plants did not contain
the NPT II gene, four of the 39 heterozygous NPT 11' plants were self pollinated in the
greenhouse. The progeny of three segregated as predicted for the native gene, three
green; one vir (154157; x2 = 0.456, df =1). As was the case for the parents, none of
the resultant F 2 progeny (n = 211) expressed detectable levels of NPT H protein by
ELISA (data not shown), nor could the NPT II gene be detected with PCR (n=20).
The segregation of selfed progeny from the fourth plant indicated that a heterozygous
green, (Xv), non-transgenic mother may have been the donor of the 12 NPT 11’ plants.
The progeny of this plant were all green. This outcome could not be explained by a
donor plant that was heterozygous for the NPT H gene (1/4 of the progeny would be

expected to be vir). Importantly, this outcome also could not be explained by

79

Figure 4. PCR products resulting from a reaction run with the two 18 base pair
primers flanking a 700 base pair region internal to the NPT H gene, and genomic
DNA from: Lanes 2-3, NPT II ELISA’“ green seedlings; Lane 4, NPT H ELISA“
virescent seedling; and Lanes 5-10, NPT H ELISA' green seedlings. Band in lanes 2-3
and 5-10 corresponds to approximately 700 base pairs based on comparison with the

Hind IH digested A DNA size standard, Lane 1.

 

81

differential movement of native and engineered genes (again 1/4 of the progeny would
be expected to be vir). This outcome, (all green progeny) only could have occurred if
this fourth plant was homozygous for the non-virescent (fl) allele. For that to
happen, it must be the offspring of a heterozygous green, (31v), non-transgenic mother
that was an accidental contaminant in the border plot. We have noted on occasion a
rare green cotyledon seedling in the virescent seed lots which are typically rouged.
Unfortunately it appears one escaped rouging to be planted in the plot border. The
various progeny we noted then would be the results of fertilizations from a mixture of
pollen including; self-pollen from the mother (which created fl Xv and W NPT H
negative progeny), pollen from the homozygous green transgenic donors (creating fl
and y_v NPT H positive progeny) and pollen from homozygous yellow, non-transgenic
border plants (creating y_v and W NPT 11 negative progeny).

The occurrence of the 27 plants in this study which were NPT H ELISA', but
were later found to be NPT H positive by PCR has implications for any study
designed to measure transgene movement. Studies which only screen for the presence
of a selectable marker (gene product), such as herbicide or antibiotic resistance, run
the risk of rmderestimating the actual levels of transgene movement due to transgene
inactivation phenomenon. In order to avoid these underestimates, the most thorough
way to analyze progeny in such studies would be to screen for the presence of the
actual transgene using PCR or Southern analysis. However, these progeny screens
would be enormously expensive and time consuming. The presence of a second
marker in our studies, (green cotyledons), allowed us to detect outcrossing in the

absence of NPT H expression. The use of such a dominant marker gene to identify

82

potential outcrossers as a primary screen, coupled with a secondary screen for actual
transgene presence in a more limited population presents a tractable way around dollar
and time constraints in such studies.

In conclusion, our data support the assumption that native and transgenes have
the same pollen-mediated gene dispersal patterns. This confirmation allows for the
confident use of a rich body of information on non-transgenic plants in making
predictions about patterns of gene dispersal via pollen in transgenic plants. Although
gene flow information is an important component relative to the establishment and
spread of transgenes in natural ecosystems, other factors must also be considered.
Specific features of the particular transgene (i.e. does it confer a selective advantage)
will strongly influence performance of the progeny, this in turn will effect the rate and
extent of establishment and spread (Williamson, 1994; Regal, 1994; Gliddon, 1994;

Linder and Schmitt, 1994; Gabriel, 1993).

83

Literature Cited
Andow, DA. 1994. Community response to transgenic plant releasersing

mathematical theory to predict effects of transgenic plants. Molecular Ecology

3 265-70.

Beck, E., G. Ludwig, EA. Auerswald, B. Reiss and H. Schaller 1982. Nucleotide
sequence and exact localization of the neomycin phosphotransferase gene from

transposon T115. Gene 19:327-336.

Bergelson, J. 1994. Changes in fecundity do not predict invasiveness: A model study

of transgenic plants. Ecology 75(1):249-252.

Colwell, R.K., E.A. Norse, D. Pimentel, F .E. Sharples and D. Simberloff 1985.

Genetic engineering in agriculture. Science 229:111-112.

Crawley, M.J. 1987. What makes a community invasible? pgs. 429-453 in
Colonization, Succession and Stability. [eds] A.J. Gray, M.J. Crawley and PI.

Edwards. Blackwell Scientific Publications, Oxford.

Crawley, M.J. 1990. the ecology of genetically engineered organisms: Assessing the
environmental risks. pgs. 133-150 in Introduction of Genetically Modified
Organisms into the Environment [eds] HA. Mooney and G. Bernardi. Scope

44 John Wiley & Sons, Chichester.

84
Crawley, M.J., RS. Hails, M. Rees, D. Kohn, and J. Buxton 1993. Ecology of

transgenic oilseed rape in natural habitats. Nature 362:620-623.

Dale, P.J. 1992. Spread of engineered genes to wild relatives. Plant Physiology

100:13-15.

Darmency, H. 1994. The impact of hybrids between genetically modified crop plants
and their related species: introgression and weediness. Molecular Ecology

3:37-40.

Dellaporta, S.L., J. Wood and JB. Hicks 1985. Maize DNA miniprep. in Malmberg,
R, J. Messing and I. Sussex [eds] Molecular Biology of Plants. pgs. 36-37.

Cold Spring Harbor Laboratory, NY.

Ellstrand, NC. 1988. Pollen as a vehicle for the escape of engineered genes? Pgs.
530-532 in J. Hodgson and AM. Sugden [eds] Planned Release of Genetically
Engineered Organisms. Trends in Biotechnology/Trends in Ecology and

Evolution Special Publication. Elsevier Publications, Cambridge UK.

Ellstrand, NC. and CA. Hoffman 1990. Hybridization as an avenue of escape for

engineered genes. BioScience 40:438-442.

85

Fang, G. and R Grumet 1993. Genetic engineering of potyvirus resistance using
constructs derived from the zucchini yellow mosaic virus coat protein gene.

Molecular Plant-Microbe Interactions 6(3):358-367.

Finnegan, J. and D. McElroy 1994. Transgene inactivation: Plants fight back!

BioTechnology 12:883-888.

Gabriel, W. 1993. Technologically modified genes in natural populations: Some
skeptical remarks on risk assessment from the view of population genetics. in
K. Wohrrnann and J. Tomiuk [eds] Transgenic Organisms: Risk Assessment of

Deliberate Release. Birkhauser, Basel

Gliddon, C. 1994. The impact of hybrids between genetically modified crop plants
and their related species: biological models and theoretical perspectives.

Molecular Ecology 3:41-44.

Handel, SN. 1982. Dynamics of gene flow in an experimental population of Cucumis

melo (Cucurbitaceae). American Journal of Botany 69(10):]538-1546.

Hokanson, S.C., R Grumet and J.F. Hancock 1996. Effect of border rows and
trap/donor ratios on pollen-mediated gene movement. (Submitted Ecological

Applications)

86

Kareiva, P., W.F. Morris and CM. Jacobi 1994. Studying and managing the risk of
cross- fertilization between transgenic crops and wild relatives. Molecular

Ecology 3:15-21.

Klinger, T., D.R Elam, and NC. Ellstrand 1991. Radish as a model system for the
study of engineered gene escape rates via crop-weed mating. Conservation

Biology 5(4):531-535.

Linder, CR. and J. Schmitt 1994. Assessing the risks of transgene escape through

time and crop-wild hybrid persistence. Molecular Ecology 3:23-30.

. 1995. Potential persistence of escaped transgenes: Performance of
transgenic, oil-modified Brassica seeds and seedlings (in press) Ecological

Applications.

Manasse, RS. 1992. Ecological risks of transgenic plants: Effects of spatial dispersion

on gene flow. Ecological Applications 2(4):431-438.

Manasse, R and P Kareiva 1991. Quantifying the spread of recombinant genes and
organismspgs. 215-231, In Ginzburg, L. [ed] Assessing Ecological Risks of

Biotechnology.Butterworth-Heinemann, Boston, MA.

87
McPartlan, BC. and P.J. Dale 1994. An assessment of gene transfer by pollen from

field-grown transgenic potatoes to non-transgenic potatoes and related species.

Transgenic Research 32216-225.

Mooney, HA. and J.A. Drake 1990. The release of genetically designed organisms in
the environment: lessons from the study of the ecology of biological invasions.
In: Introduction of Genetically Modified Organisms into the Environment [eds]
Mooney, H.A., G. Bemardi; Scientific Committee on Problems of the

Environment 44), pp. 117-127. John Wiley and Sons, Chichester.

Morris, W.F., P.M. Kareiva and PL. Raymer 1994. Do barren zones and pollen traps
reduce gene escape from transgenic crops? Ecological Applications 4(1):159-

165.

Nugent, PE. 1987. C879-Jl and 42 virescent mutant muskmelon breeding lines.

HortScience 22(2):333-335.

Purrington, CB. and J. Bergelson 1995. Assessing weediness of transgenic craps:

Industry plays plant ecologist. Trends in Ecology and Evolution 10(8):340-342.

Raybould, AF. and A.J. Gray 1994. Will hybrids of genetically modified crops

invade natural communities? Trends in Ecology and Evolution 9(3):85-89.

88

Regal, P.J. 1994. Scientific principles for ecologically based risk assessment of

transgenic organisms. Molecular Ecology 325-13

Rissler, J. and M. Mellon 1993. Perils amidst the promise: Ecological risks of
transgenic crops in a global market. Union of Concerned Scientists,

Cambridge, MA.

Rissler, J. and M. Mellon 1995. What's coming to market? The Gene Exchange

6(2&3):8-9

Scheffler, J.A., R Parkinson, and P.J. Dale 1993. Frequency and distance of pollen
dispersal from transgenic oilseed rape (Brassica napus). Transgenic Research

2:356-364.

Williamson, M. 1994. Community response to transgenic plant release: predictions
from British experience of invasive plants and feral crop plants. Molecular

Ecology 3:75-79.

CONCLUSIONS

89

90

CONCLUSIONS

The objectives of this project were to: 1) Investigate the efficacy of border
rows as a strategy to limit the pollen-mediated escape of transgenes from crop
plantings and 2) Determine whether native and transgenes have the same pollen-
mediated dispersal patterns.

Border rows have been incorporated into many of the field tests of genetically
engineered creps to satisfy requirements for restricting the escape of engineered genes.
Although a limited number of studies have investigated the movement of transgenes
into a contiguous border of non-transgenic cogeners, no systematic study of the
influence of border rows on pollen-mediated long distance gene movement to isolated
patches of co-geners has been reported. Research along these lines is important
because such an experimental design closely approximates the situation where patches
of crop wild relatives surround commercial crop plantings.

Our results demonstrate that border rows can have a significant influence on
long distance gene movement to discontiguous satellite plots. As the trap plant to
donor plant ratio increased from 0 - 131, there was a significant decrease in long
distance gene movement. In the plots with a ratio of 131 we saw no long distance
gene movement into the satellites.

Despite the significant treatment effect observed in these experiments, the

9]

border row strategy has several inherent weaknesses. The principle problem is
economic in nature. Even our smallest trap/donor ratio which allowed on average
0.61% outcrossing to the satellite plots, would necessitate the planting of 11.6 acres of
border plants for every acre of transgenics. This does not appear to be an
economically viable strategy for a grower. From an environmental perspective, some
might find the less than perfect protection afforded by the strategy objectionable.

The other significant issue raised by this study concerns environmental
variation. Year to year and site to site variation in long distance gene movement
suggests that broad predictions regarding the influence of border rows as a
containment strategy can not be made. The 38% outcrossing rate observed in one of
the satellites is an example of a ”low frequency, high magnitude" type event some
environmentalists fear could result from the unregulated commercialization of
genetically engineered crops.

Results from our experiments compel us to conclude that regardless of the
border row strategy employed, some level of escape of engineered genes from
agronomic-scale plantings appears inevitable.

Much of the response to issues regarding the potential risks of genetically
engineered crops has been based on along standing body of information generated
from research with non-transgenic organisms. Application of such information to risk
assessment for transgenic plant issues necessitates the assumption that there is no
inherent difference in pollen-mediated dispersal patterns for native and transgenes.
While this assumption is probably valid, it remains untested. To our knowledge, no

reports exist where native and transgene dispersal patterns are compared.

92

Long distance dispersal patterns of the native and transgene in this study were
identical. At no time did we see movement of the transgene without a concomittant
movement of the native gene or vice-versa Regarding short distance movement into
the plot borders, the dispersal patterns for the two genes were nearly identical. The
slight difference observed for short distance dispersal, (0.30%), was accounted for by
the unfortunate occurrence of a contaminant plant in one of the plots. Aside from this,
there was no unaccountable difference in dispersal patterns between the two genes.
These results suggest that the application of information generated from research on
non-transgenic organsims will be legitimate and useful.

Two important cautionary issues must be raised at this point. The first is the
phenomenon of transgene inactivation. Transgene inactivation is basically the lack of
expression of the transgene itself and/or endogenous genes, apparently resulting from
the insertion of DNA into the genome via the transformation process. The
phenomenon has resulted in some unpredicted phenotypes which in some cases
appeared to be the result of abberant transmission patterns. Discovery of 27 NPT H
ELISA' PCR+ seedlings originating from the Sandhill plot in these experiments
(presumably a result of transgene inactivation) raises important implications for any
similar experiments evaluating the risk of pollen-mediated escape of engineered genes.
Studies designed to look only for presence of a transgene product such as herbicide or
antibiotic resistance, as a measure of transgene movement, run the risk of
underestimating the actual amount of gene movement due to the transgene inactivation
phenomenon . A true measure of transgene movement can only be gained through

screening for presence of the actual transgene. In the case of our experiments, the

93

difference in the two measures would have been nearly one percent.

The second caution does not arise directly as a result of this research, nor is it
new. The fact that the dispersal patterns for native and transgenes in this study were
the same does not suggest that the transgenes will not have a unique impact in the
natural environment. It is important that the adaptive value of each gene be evaluated
individually. Each gene has the potential to impart on its new host a particular fitness
value which could make that transgenic plant more or less competitive relative to its
neighbors. In fact, with the ability to move genes between genera, families and even
kingdoms, scientists have the ability to impart on plants fitness traits and thus selective
potentials never seen before in the plant community! Persistance and spread of
transgenes in the environment are not parameters concerning risk assessment about
which we can necessarily make assumptions. Such decisions will to a large degree be
made on a gene/crop by gene/crop basis.

Given the high likelihood of some level of escape of transgenes from
commercial scale plantings, the highest priority for future risk assessment research
should be in the area of establishment and spread of transgenes in the environment.
Specifically, can the risk of escape of transgenes into the environment be mitigated at
this level? The focus of such research should be on the influence of the selective
value imparted by a particular transgene on the likelihood of its persisting and
spreading in the environment. The advent of biotechnology presents an excellent
opportunity to "build" plants with a quantifiable fitness value in a given environment,
i.e. herbicide resistance under certain herbicide spray regimes. Experiments could be

designed to test the persistance of and rate of spread of transgenes under selection

94

conditions which reduced the fitness of non-transgenic controls by 25, 50 or 75%.
Such direct, quantitative selection experiments would provide valuable information to
both the risk assessment and evolutionary biology communities.

With such background information in hand, predictions regarding the potential
risks posed by specific gene/crop combinations might be made. At the least, such
information would allow more informed decisions to be made regarding which
gene/crop combinations merit the most rigorous oversight and investigation.

Such prioritising is critical as agricultural biotechnology firms seek to begin
gaining some profit from their investments. Strict regulations over all products of
biotechnology will result in one of two eventualities. Lack of profits could prompt
biotech firms to discontinue biotech pursuits. Alternatively, strict across the board
regulation could result in a move to end all regulation of biotech products. Neither
scenario is acceptable. Although the present day products of agricultural
biotechnology are not producing ”gardens in the desert", at some time in the future
they will in all likelihood do that and more. To acheive that end the industry needs to
be allowed to prosper. However, we can not allow progress at any cost. Just as
certainly as there are products of biotechnology which are overregulated today, there
are products in the pipeline which will merit strict oversight.

With all said and done, risk assessment of genetically engineered crops will to
a large extent be relegated to a case by case basis. Certain crop/gene combinations
may well be deemed unallowable, while others will necessitate as little oversight as

present day conventionally bred crop varieties.

APPENDICES

95

085... 88 300.08.

.2800: 50 A0885: 3 .5 0.088% .5828 088 .80 058 0.3 08005 808000.?

.5880 803 :85— 8 0058 .0. A 588 =< A895 88.500 3 000.. .58 5>0 55.5 88.. Ba 20 80888 8:?

6.5 88850 “a 000» .88 55 555.6 80508 50 0098006 803 5a 0.8 80:88 526EN

.0830 .800— 88 05 .0 008300 203 358000 088803

.030.— 5053 80 K .30.. 05 859» 8 n. 58% 203 35:— 808008 5.— 353 m .000 5a 3550.0 m .000.— .58 .Dmum.

 

 

 

20.0 «0 32.0 a 80.08 80040 8.040 80.040

2 0.0 I 00 E 2 208.00 0.0
@0040 28.0.0 20.0.0 808 80.08 30.08

I 2 0.0. 0.0 _.~_ 3 8.8.00 3
28.08 5.040 30.08 E08 8.10 2008

E .2 3 0.0 0.2 0.0 258.3 3
30.040 28.040 0.0.08 308 9.3 80.08

0.0 2 0.0 0.0 a: 2 288.00 2
5.08 E08 3.08 3.08 $0.08 3.00

2 0._ 3. 0.0 02 an 0 0.0
83 N0 08 00 .20 8 80 00 “20 8 8.3 00 088020

«0532803 05—88038". 0 ”05—88030—0— .p

 

«38888 83 6:02.80 .0088 .8808; .2828. as» $85.08 see a 80

.3835 880083. 090.53 5:855 05850-00: .80 8058008 88 .05 .8305 8850 08080.82 8 0058.08 A 8059?.

96

00.053 50. 32.3.

.0550: 5a 30535: R .5 20050.6 2580055 03.00 .80 2:05 20 358:8 25.5000?

2:0 .08 00308 053 .35 mm 00 0500

03m .595— 5 00:05 N A :8 :0 053 m<n 00 0.0 3:000 tam @0005 050:5... ”a 000: B8 5>0 00w05>0 0:03 50 000 050:5: 00:":
A085 0:00:50 “3 090.. .58 5>0 00005.30 0:08.09: 50 00w05>0 0:03 50 0:0 052:5: 530...“N

.03.... .800— 58 0.0 5 000.50 053 0:53.000

0053.8; .0300 .5035; :5 Z. .38 05 :003 8 n. 000000 053 3:03 000.5005 50 3:0:— m .00.. 50 3.5.5005 n .000: 58 .Dmuyr

 

 

 

 

3...... 00 8.0 «v 5.0.0 8.30 83.0 83.0
2 0.0 0.0 0.0 3 S 25800 0.00
5.0.0 8.90 20.20 5.30 32.0.0 82.0.0
0... 0.0 3 2. .3 3 23850 0.:
8.00 823 80.0.3 8.000 33.0 5.30
0.0 :0 0.0 0.0 3 3 238.00 00
3.0.0.0 823 832 80.30 5.00 823
in. 2 E .3 3 3 828.00 E.
32.0 $.30 $3.0 5.0.0 3.3 $0.00
00 2 Z 3 Na 3 0 0.0
.20 a .20 3 .20 3. .20 mm .20 3. 2:5 mm 05.5020
mfigem .535320 0 002020320 .0

 

052.20... 8.: 2282.030 sane 000.02. .2805. 55 25.580 020... 2 2.0
.3838 530005? bong 5280000 0053:0075: .80 €002.08: .08 0:0 5305 $00000 030090030.— 5 5000008 .N 50:00.?